AD-A089 606 BATTELLE COLUMBUS LABS OH F/6 11/1POLYMER RESEARCH IN RAPID RUNWAY REPAIR MATERIALS. (U)

UNLSIIDNOV 79 M LUTTINBER. C W KISTLER. H M BROTTA F06635-79-C 0040W SIFIED AFESC/ESL-TR-79-43 NA

illulllllllluIIIIIIIIIIIIIIEiIuIIIIIIIIIu-EEIEEEEIIEEEIIIIIIIIIIIIu

11111 11.51 0 I12

11111"2----5 II1.4 1i1ll.

MICROCOPY RESOLUTION TEST CHART

..M-71-7S-43

I LEVEVLPOLYMER RESEARCH Il RAPID RUNWAY

SREPAIR MATERIALS

let

BATTELLE COLUMBUS LABORATORIES

COLUMBUS, OHIO 43201

NOVEMBER 1979

FINAL REPORTJAUARY 1979- OCTOBER 1979

/ ( DTICELEECTE-

APPROVED FOR PUBLIC RELEASE; DISTR!U1TION UNi D

ANINE EUG I A SERVE LA oRAOR9 AN L49 2

Please do not request copies of this report f .

HQ AFESC/RD (Engineering and Services LabortoryL).

Additional copies may be purchased from:

National Technical Information Service

5285 Port Royal Road

Springfield, Virginia 22161

Federal Government agencies and their contractors.

registered with Defense Technical Information Center

should direct requests for copies of this report to:-,

Defense Technical Information Center

Cameron Station

Alexandria, Virginia 22314

MAN"-

UnclassifiedSEE UNITYF CL A%%1 IC ATSItN Of TMI' .IAr.. (WI.... 0 *, . v

~ REPORT DOCUMENTATION PAGE RE~AItDW I I ONS:

4TITLE (and Subtitle) --

POLYMER RESEARCH IN RAPID RUNWAY REPAIR HATERIS Jan y.-Ocn~*96 9P.ERFORING ONGc. IEiFlT ..... L

7. AUTHOR(j antreqLuttinger 5r~

* ~ lery H. rotF 0'635-79-C-4*,S /Rcad G/Sinclair ____________

9. PERFOR~ttlfF t In tktSK

Battelle Columbus Laboratories V'ANNA "MORK UNIT NUMBER$

505 King Avenue TN~7O 1 '~Columbus, OH 43201 __________

It. CONTROLLIN4G OFFICE NAME AND ADDRESS

) Air Force Engineering & services Center & No! 7 -i9Tyndall Air Force Base, Florida 32403 92*.-~ r E

24. MONITORING AGENCY NAME It ADORESS(Ut different from Controlling Office) IS. SECURtITY CLASS. (of this. report)

Unclassified

0. ECLSSIFICAiTIjOwDWNGRAINGSCHEDULE

ii. DISTRIBuTioN STATEMENT (of this Ropoii)

Approved for public release; distribution unlimited

17. LOISTRIDUT1014 STATEMENT (of ithe abstrt entered In Block 20. #1 different from R~pot

18. SUPPLEMENTARY NOTES

Availability of this report is specified on verso of front cover.

19. KEY WORDS (Continue an reverse side if necesseary and Identify by block numbofa)

Bomb Damage Repair Epoxy Resins Rapid Runway RepairCivil Engineering Pavement Repair SilanesConcrete Repair Polymer Titana%Coupling Agents Polymer Concrete

*%.ABSTRACT (Continue an revere. dide If necesary end Identit or block nun.ber)

'Umvisosiytwo-component e~qw-resins wer ormulated for a/irless sprayapplication over quartz or delomite aggr~g~s Th omuai eete ofull evaluation was based on mrag1,curing systems. Tri 1ioa scry-late monomers were used in soIMelebnulat ions as modifiers. T~e resultingpolymer cotrete set up w 5hi63 to 4 minutes after mixing a temperaturesaround 73* F. Good c"p within 1/2 hour of maixing can be oftained in vet en-vironments down to 50 C and In dry environments down to -250 C. Good adhesionto vet aggregate. require. the use of coupling agents, organofunctional *ilanes

DD I jll7 1473 EDiTWoN orFI, NOV GS 15 OIISOLETE Unclassifiled4 ~~~SCCURITY C1. A..IrICATI"14 UV I fll PA01 (Pton oueIta~~

Unclassified

seCUpIoY CLASSIrICAYIOId OF TMIS PAGE(When Dlate M -ed)

being preferred. Good bonding to asphalt and Portland cement concrete and gcodwear characteristics were demonstrated. Flexural strength properties are satis-factory after cool down both under dry and wet application conditions. While thepolymer concrete is hot due to the exotherm of the curing reaction, flexuralstrength properties are low.

UnclassifiedII C UIITY CLA',%IFICAIIO4 or Til;- SACS.

PREFACE

This report was prepared by Battelle Columbus Laboratories,505 King Avenue, Columbus, Ohio, 43201, for the Air Force Engineering andServices Center (formerly Air Force Civil and Environmental EngineeringDevelopment Office, Air Force Systems Command),Tyndall Air Force Base,Florida, 32403, under Contract No. F08635-79-C-0040 Mr. Richard G. Sinclairwas project manager; Hr. Manfred Luttinger was principal investigator;and Mr. Charles W. Kistler, Jr., and Mr. Henry M. Grotta were coinvestigators.D. Mangaraj contributed to the analysis of factors contributing to heatbuild-up and dissipation during curing of epoxy resins. This reportsummarizes work done between January and October 1979. Captain MichaelT. McNerney (AFESC) was project officer.

This report discussed the use of many.name brand epoxy resinsand components. This report does not constitute an endorsement of theseproducts by the Air Force nor can it be used to advertise the products.

This report has been reviewed by the Public Affairs Office (PA)and is releaseable to the National Technical Information Service (NTIS).At NTIS it will be available to the general public including foreignnations.

This report has been reviewed and is pr ed for publication.

7%IftMICHAEL T. MCNERNEY, Cant, USAF ROBERT E. BOYE2LJ Col., USAFProject Officer Chief, Engineer Research Division

GEORGE D. BALLENTINE, Lt Col, 11sAWDirector, Engineering and Services Laboratory

Ac essi0U for

DDC SLB

Justifc iat ion

Di stributi

(The reverse of this page is blank.) Ama&ilud/or

Dist sped

TABII', OF 'N'ICI:N'I'S

Section TJ t I

I. INTRODUCTION ....... I

II. OBJECIVE ............. .......................... 2

III. SUMMARY .............. .......................... 3

3.1 Resin and Curing System Evaluation ..... ........... 33.2 Coupling Agent Evaluation ........ ................ 43.3 Performance Evaluation of Selected Resin Systems . . .. 4

IV. RESEARCH APPROACH ........... ...................... 6

V. EXPERIMENTAL PROGRAM .......... ..................... 8

5.1 Resin Formulation and Screening ...... ............. 85.1.1 Testing Method ......... .................. 85.1.2 Epoxy Resin Selection ....... .............. 85.1.3 Polyamide Cures ...... ................. ... 105.1.4 Amine Cures ....... ................... . I.115.1.5 Mercaptan Cures ...... ................. ... 135.1.6 Mercaptan/Acrylic Monomer Cures .. ........ ... 155.1.7 Polyamide and Amine/Mercaptan Cures . ....... . 165.1.8 Lewis Acid Cures ...... ................. ... 16

5.2 Selection of Coupling Agents for Use inPolymer Concrete ........ ................... ... 185.2.1 Union Carbide Silanes .... ............ ... 19

5.3 Resin Bonding to Aggregates ...... ............. ... 205.3.1 General Procedure .... .............. .... 205.3.2 Substrate Selection ..... ............... ... 225.3.3 Results of Bonding Study .... ............. ... 22

5.4 Adhesive Strength to Asphalt and Concrete .......... ... 255.4.1 Substrate Selection . . . ............ 255.4.2 Results of Adhesion Tests ... ............ ... 25

5.5 Flexural Strength of Polymer Concrete ... .......... ... 285.5.1 Aggregate Selection/Characterization ........ ... 285.5.2 Specimen Preparation Procedure ... .......... ... 305.5.3 Results of the Flexural Strength Tests ....... ... 30

5.6 Durability ......... ....................... ... 355.6.1 Background ........ .................... ... 355.6.2 Test Procedure ....... .................. ... 355.6.3 Control Materials ..... ................ ... 375.6.4 Results ........ ..................... ... 37

5.7 Storage Stability of Resin Systems ... ........... ... 39

VI. EQUIPMENT REQUIREMENTS ........ .................... ... 41

VII. CONCLUSIONS AND RECOMMENDATIONS ...... ............. ... 43

7.1 Discussion of Theoretical Considerations . ........ ... 447.1.1 Rate of Polymerization (Rp) ... ........... . 447.1.2 Heat of Polymerization (AHp) ... ........... ... 477.1.3 Temperature Profile ..... ............... ... 477.1.4 Heat Transfer (Rt) ...... ................ ... 477.1.5 Temperature Rigidity ..... ............... ... 47

iii

TABLE OF CONTENTS (CONCLUDED)

Appendix Title Page

A. SUMMARY OF TABLES OF EPOXY FORMULATIONS ... ........... ... 49

B. SELECTED POLYMER CONCRETE LITERATURE REVIEW .. ......... ... 77

C. PRICE, HANDLING, AND SAFETY INFORMATION OFFORMULATION COMPONENTS .... .... .................... ... 80

LIST OF FIGURES

Figure Title Page

1. Adhesively Bonded Cross-Block Tensile Test Specimenin Testing Rig (Each block lxlx2 inches) ... ........... ... 21

2. Effect of Aggregate Diameter on Permeability .. ......... ... 29

3. Application of Resin with Airless Spray Equipment onAggregate-Loaded Test Trays ...... ................. ... 32

4. Flexural Testing of Beams ...... .................. ... 32

5. Device for Durability Tests ...... ................. ... 36

6. Asphalt Concrete Sample After Durability Testing ........ ... 36

7. Abrasive Wear of Concrete Materials Versus Time . ....... . 38

ivi

LIST OF TABLES

Table Title Page

1. Adhesive Performance of Epon 828-Based Epoxy ResinFormulations to Untreated Wet Silica and LimestoneSurfaces ........... ........................... ... 23

2. Adhesive Bond Strengths (Psi) of Epoxy Compositionsto Treated and Untreated Silica and LimestoneSubstrates Under Wet Conditions at 730 F ... ........... ... 24

3. Adhesive Bond Strengths (Psi) of Epoxy Compositionsto Treated and Untreated Silica and LimestoneSubstrates Under Wet Conditions at 40Pto 450 F .......... .26

4. Adhesive Bond Strength (Psi) of Epoxy Composition71-3 to Asphalt and Portland Cement Concrete .. ......... ... 27

5. Properties of Aggregates Evaluated ..... .............. ... 31

6. Results of Flexural Strength Measurements ... .......... .34

7. Accelerated Aging Test of Resin Components at 60* C ...... .. 40

8. Effect of Substituents on Tertiary Amine on Reactivityof DGEBA ............ .......................... .46

9. Heat of Polymerization of DGEBA With VariousCuring Agents .......... ........................ .48

A-1. Polyamide Cured Formulations Based on Epon 812 andApogen 101 Epoxy Resins ....... ................... ... 50

A-2. Polyamide Cured Formulations Based on Apogen 101and DER 732 Epoxy Resins ....... ................... ... 51

A-3. Polyamide Cured Formulations Based on DEN 431Epoxy Novolac Resin ........ ..................... ... 52

A-4. Polyamide Cured Formulations Based on Epon 828Epoxy Resin .......... ......................... ... 53

A-5. Amine Cured Formulations Based on Araldife 509,

Capcure WR and Epon 828 Epoxy Resins .... ............. ... 54

A-6. Amine Cured Formulations Using Salicrylic Acid/

Amine Adducts as Curing Agents ..... ............... ... 55

A-7. Amine Cured Formulations Based on Epi-Rez/Epi-CureCombinations ........... ........................ . 56

v

LIST OF TABLES (CONCLUDED)

Table Title Page

A-8. Amine Cured Formulations Based on Combinations ofEpon 828 and Various Ancamine Curing Agents. .......... 57

A-9. Mercaptan Cured Formulations Based on Epon 812Epoxy Resin and Mixtures of Epon 812 with Epon828 and Apogen 101 .. ...................... 59

A-10. Mercaptan Cured Formulations Based on Epon 828Epoxy Resin ...... .................... 60

A-li. Mercaptan Cured Formulations Based on Low ViscosityEpoxy Resins and Mixtures. ................... 61

A-12. Mercaptan Cured Formulations Based on Capcure WRand Epon 828 Epoxy Resins. ................... 63

A-13. Mercaptan Cured Formulations Based on Araldite 509Epoxy Resin ........ .................. 64

A-14. Mercaptan Cured Formulations Based on Epon 828 EpoxyResin and Low Viscosity Mercaptate Esters. ........... 66

A-15. Mercaptan Cured Formulations Based on Apogen 101,D.E.N. 431 and Epi-Rez As The Main Epoxy ResinComponents. ......... ................. 67

A-16. Mercaptan Cured Formulations Based on Epon 828/Heloxy 69 Epoxy Resin Mixtures. ....... ......... 68

-'A-17. Evaluation of Polyfunctional Acrylic Monomer AsD~iluents in Mercaptan Cured Epoxy Formulations .. ........69

A-18. Additional Formulations With Polyfunctional AcrylicMonomers As Diluents in Mercaptan Cured EpoxyFormulations. ....... .. .................. 70

A-19. The Use of Apogen 256 and Versamid 125 in the MixedPolyamide/Mercaptan Cure of Epoxy Compositions .. ........71

A-20. The Use of Capcure 38 in the Mixed Polyamide/Mer-captan Cure of Epoxy Compositions ........ ....... 72

A-21. DP116 Cure of Araldite 509 Epoxy Resin. ....... ..... 73

A-22. Resicure 30 Cure of Epon 828 Epoxy Resin. ............ 74

A-23. Resicure 30 Cure of Various Epoxy and Epoxy/Polyamide Resin Compositions. ............. ....

A-24. Lewis Acid Cures of Cycloaliphatic Epoxides ........... 76

vi

SECTION I

INTRODUCTION

The need for rapid runway repair to restore operational capabilityfollowing an enemy attack has been well recognized, and provisions for suchcontingencies in the NATO theatre are in effect in accordance with AFR 93-2.Repair methods are based on the use of AM-2 matting anchored over backfilledcraters and have been designed primarily for damage caused by 750-lb bombsleaving large bomb craters of about 50-foot diameter.

Diversification of repair methods is required to meet the threat ofnew weapons technology which has the capability of producing a large numberof smaller craters (about 20-foot diameter). Existing repair methods areinadequate under the changed circumstances, leaving a flight surface of ex-cessive roughness. Inorganic binders which were evaluated were also un-acceptable from a performance standpoint. Several research programs invest-igating the use of organic resin binders reported promising results. How-ever, shortcomings such as limited shelf life, complexity of application,moisture sensitivity, and insufficient adhesion still remain in the mostpromising candidates which include acrylics, polyesters, and epoxy resins.

The contractor approach has involved a systematic study of epoxyresin technology to identify polymer-concrete systems for use in rapid runwayrepair under the extreme conditions of temperature and moisture prevalent inthe North Atlantic Treaty Organization (NATO) theatre. The present programhas focused on the following three technical problems that must be overcometo develop a satisfactory repair system:

(1) Adhesion to wet aggregates.(2) Cure rate control for low temperature service.(3) Control of resin viscosity.

In addition, resin bonding to asphalt and Portland cement concrete was stud-ied, durability and flexural strength properties of the polymer-concrete weredetermined, and accelerated shelf life stability of the resin components wasevaluated.

a1

SECTION II

OBJECTIVE

This research and development program had the objective of invest-igating a broad range of epoxy resin and curing systems for development of a

satisfactory repair method for bomb-damaged runways. An optimized resin-aggregate system was sought for use under a wide range of environmental tem-perature and moisture conditions, particularly as may be encountered in theNATO theatre. To this end, it was desired to accommodate temperature ex-tremes of -25* to 1250 F under wet and dry conditions. Flexural strengths of400 psi needed to be developed in the resin-aggregate system. Bond strengthsof at least 40 psi were required between resin and either Portland cementconcrete or asphalt. The cure time of the resin-aggregate cap over the re-paired crater should be held to within 30 minutes after application. Minimumshelf life of the resin system of one year at 720 F was required.

2

SECTION III

SUMMARY

3.1 Resin and Curing System Evaluation

The present program included an evaluation of a broad range of

epoxy resins and curing systems. Initially, specialty resins were investig-ated because of cheir reportedly higher reactivity in order to meet the rapidcure requirements of the program. It was eventually demonstrated, however,that the DGEBA (diglycidyl ether of bisphenol-A) type resins could also meetthe rapid cure requirements, both under cold and wet conditions, provided theappropriate curing system was utilized. In short, the curing system was therate determining factor, and the particular epoxy resin usually played onlyan insignificant role. The viscosity and cost penalties imposed by most ofthe specialty epoxy resins could therefore not be justified.

Another important consideration in the selection of epoxy resinswas their viscosity, since both the spray application and the penetrationthrough potentially cold or wet aggregates require highly fluid resin com-binations. A number of low-viscosity resins were included, therefore, in thestudy, as well as a variety of diluents. The latter included such nonreact-ive diluents as hydrocarbon resins as well as reactive diluents of both theepoxy-functional and nonepoxy variety. As might be expected, the bestviscosity reduction with minimum compromise of hardness and strength pro-perties was achieved with epoxy-functional (especially difunctional)diluents.

The resin system selected for full-scale evaluation was based on acombination of a standard DGEBA resin of 11,000 to 15,000 cps viscosity (Epon828) and a resorcinol-di-glycidyl ether diluent (Heloxy 69). Also of p. t-ential interest for future study was a compounded DGEBA mixture with EpoxideNo. 7 (Araldite 509). Three low-viscosity epoxy resins of proprietary com-

position were Capcure WR Epoxide and Epi-Rez 5027 and 50727. The latter tworesins, however, were limited in their reactivity to amino-functional curingsystems.

A very effective diluent type for admixture with epoxy resins wasfound to be polyfunctional acrylic monomers which combined low viscosity andnonreactivity with the epoxides with fast reaction rates with mercaptan cur-ing agents. Acrylate diluents are limited, however, for use with this curingsystem and constitute a dual cure: acrylate-mercaptan and epoxy-mercaptan.An interpenetrating network, with the mercaptan resin forming the bridgebetween the epoxy and the acrylic constituents, may be postulated.The most effective acrylate monomer which was selected for full-scale eval-

uation was trimethylolpropane triacrylate.

A variety of curing agents were evaluated in the course of thisprogram including several examples of polyamides, amido-amines, amines, mer-captans, and Lewis acids. In addition, several combinations of these curingagents were evaluated. The screening criteria used for this evaluation were(1) ambient temperature cure rate and hardness, (2) cure rates under cold andwet conditions, and (3) considerations relating to field application such asviscosity and mix ratio of epoxy resin and curing agent. An overview of theresults of this screening study is given in the following tabulation.

3

Fast EqualFast Cure Rates Development of Low Mix

Curing Agents Ambient Cold Wet Hardness Viscosity Ratio

Polyamides - -- + - - ++

Amines andAmido-Amines ++ -+ - ++ - -

Mercaptans -4++ ++ ++ + -+ +4+

Mercaptans/Acrylics +++ 4-- -+ -+ -+ +++Mercaptans/Amines

and Amides 4- + -+ ++ -+ ++Lewis Acids 4+ ++ -+ + + --

- poor performance or disadvantage.+ good performance or advantage.

Polyamides were eliminated because of their relatively slow curerates. Amines and amido-amines did not usually perform well under very coldor wet conditions and had the disadvantage of unequal mix ratios. The bestoverall results were obtained with mercaptans and mercaptanfacrylic monomercombinations. The designation -4 in the above tabulation indicates widevariations between different formulations. Combinations of mercaptan resinsand amines or amides lagged behind the all-mercaptan cure system with respectto cure rate and viscosity considerations. Some of the Lewis acid cures ap-peared quite promising, but adhesion under wet conditions was a major problemand large disparities in mix ratios was a disadvantage.

3.2 Coupling Agent Evaluation

It was shown during the adhesion studies of resin systems to wetsilica and limestone aggregates that the use of coupling agents was essent-ial. Coupling agents were employed as a pretreatment for the aggregates,which is believed to be the most effective method of use, given the rapidgellation of the curing system. Several silanes with varying Organo-functional reactive groups and a titanate coupling agent were investigated.A mercapto-functional silane was selected for full-scale evaluation in keep-ing with the choice of curing system. Addition of low levels of silanecoupling agent directly to the resin system was found to aid the bonding towet asphalt and Portland cement concrete.

3.3 Performance Evaluation of Selected Resin Systems

The two mercaptan-cured resin systems were subjected to more ex-

tensive performance evaluation. These systems performed well during sprayapplication and in fast-curing performance uuder ambient and adverse temper-ature conditions and with wet aggregates. Adhesion tests to asphalt andPortland cement concrete were also quite satisfactory, far exceeding theperformance requirement of 40 psi both under dry and wet conditions, evenwithout the aid of coupling agents. Excellent results were also obtained inthe durability tests. The polymer-concrete was shown to be far superior in

4

wear resistance to the traditional concrete surfaces generallyencountered.

Flexural strength data were obtained under a variety of conditionsof temperature and dryness or wetness. The short-term flexural strengths werequite disappointing, generally falling in the range of 200 to 400 psi. Itwas shown that these relatively low values were attributable to the hightemperature of the polymer concrete at the time of testing which was causedby the high exotherm of the reaction. Flexural strength data after cooldown(tested at 24 hours) were about 1500 psi under dry conditions and 1200 psiunder wet conditions. Similarly high flexural strength values were also ob-tained within I hour of application when the test beam was briefly quenchedin ice water. Under application conditions employing cold aggregates, sub-stantially higher flexural strengths were obtained (in the 400 to 600 psirange) than under ambient conditions as a result of the heat-sink effect ofthe aggregates.

Since storage stability tests could only be started after the resinselection for full-scale evaluation had been made, accelerated aging condi-tions were used to obtain short term data. Under these severe, acceleratedconditions (600 C), considerable viscosity drift with time was encounteredwith several components. The component containing Mercaptate Q-43 ester alsoexhibited strong discoloration in the presence of oxygen. It will be nec-essary to establish whether similar observations hold true under more moder-ate storage conditions, and perhaps a switch to more stable components willbe required.

5

SECTION IV

RESEARCH APPROACH

The review of the literature had indicated that many approaches tothe rapid runway repair problem have been investigated with mixed results.The use of epoxy resin binder has been identified as among the more promising

system by several investigators. The present program concentrated exclus-ively on epoxy compositions and investigated a variety of resin and curingsystem as part of a broad formulation study.

The overall approach of the program was to concentrate on those enduse requirements that apparently were the most serious obstacles to the ob-jective of rapid runway repair of bomb damage. These include:

(1) Rapid gel time(2) Low temperature curing(3) Curing under wet conditions(4) Adhesion to wet aggregates(5) Low resin viscosity for application purposes.

For a resin system to have any chance to cure at a low temperaturesuch as -2 50 C, it was felt that a rapid gel time at ambient conditions was aminimum requirement. Laboratory effort was therefore devoted extensively toscreening of the ambient gel time of two-component resin systems in the ab-sence of aggregates. Hardness measurements were routinely monitored as anapproximate indication of the strength properties that may be potentiallyachievable. While such a correlation does not hold across different polymersystems, in the highly crosslinked, tough epoxy resins an approximation be-tween potential strength properties and hardness or rubberyness can be made.

The most successful candidates from the above screening programwere then evaluated under cold and wet conditions. Most of the formulationswere deficient in one or the other of these tests. Any resin system thatshowed promise under these severe curing conditions was suitable for opti-mization and further evaluation with regard to adhesive and strength char-acteristics.

Another major consideration for the successful utilization of aresin system was its viscosity characteristics. Low viscosity is essentialfor spray application and for penetration into the aggregate mix at low tem-peratures. Low viscosity also tends to aid the wetting of the aggregateswhich is a major requirement for the development of good adhesion.

While epoxy resins as a class are generally recognized for theirgood adhesion to a broad range of inorganic substrates, and while some epoxyformulations have been specifically formulated for adhesion to wet concrete,it was nevertheless considered advisable to investigate the use of couplingagents to enhance the bonding of resin to aggregates. Two methods of uti-lizing coupling agents are widely practiced:

(1) Pretreatment of the bonding surface.(2) Addition of the coupling agent to the resin.

6

The latter method is frequently employed because of its simplicity and be-cause it saves the extra pretreatment step. However, higher adhesivestrengths (or reinforcement in the case of filled plastics) are usually ob-tainable by pretreatment methods. For the present application, pretreatmentis considered highly desirable because the time for migration of the couplingagent to the interface is very short due to the fast gel time of the resin,and because the presence of coupling agent on the aggregates could aid wet-ting of the surface by the resin, especially under wet applicationconditions.

In order to achieve the speed of the repair required by the cir-cumstances, airless spray application of the two-component resin system isconsidered the most feasible approach. This permits the use of highly re-active systems with fast gel times, because the reactive components are notin contact until they are mixed in the spray gun. The only open time orfluidity demanded by these conditions is the time required for the resinmixture to penetrate through the aggregate to the desired depth. The latter(flow-depth) can be limited by the placing of a barrier to resin flow (asheet of nonwoven fabric, fine aggregate, or sand, etc.) at the time thecrater is backfilled.

Spray application has the further advantage of permitting theheating of the resin components (if the appropriate spray equipment isavailable) to compensate for any prevailing low temperature conditions. Inthis manner, flow into the cold aggregate can be enhanced, and the cure ratecan be boosted by giving the exothermic reaction an opportunity to start be-fore it can get quenched by the heat sink represented by the cold aggregate.

7

SECTION V

EXPERIMENTAL PROGRAM

5.1 Resin Formulation and Screening

To fulfill the objectives set for this program, it was necessary toinvestigate a large number of combinations of epoxy resins, curing agents andaccelerators in order to identify leads for compositions that could be curedboth at low temperatures and under wet conditions. Resin systems with fast

gel times under dry, ambient conditions were considered good candidates forfurther testing.

5.1.1 Testing Method

Screening of the cure rates of epoxy resin systems was conducted ondraw downs of thick films (approximately 60 to 70 mils) on glass plates andon castings of approximately 40 g samples in aluminum weighing dishes (about

2-1/2 inches in diameter, 1/2 inch deep).

The progress of the cure was followed by monitoring the tack-freetime and later the Shore hardness of both the films and the castings. Not-iceable differences were observed due to the mass effect and correspondinglyhigher exotherm obtained in the more compact castings. The latter probablymore closely simulate the actual field conditions. However, both films andcastings continued to be tested throughout the program, because the differ-ences in sensitivity to the mass effect exhibited by various formulations wasa valuable clue for comparing their exotherming behavior. It is expectedthat, in the field, the exotherm and overall cure rate will depend on theapplication temperature of the resin system, the ambient temperature of theaggregates (heat sink), the heat transfer rate to the aggregates, and the

size of the interstitial spaces between the aggregates or void volume.

5.1.2 Epoxy Resin Selection

In the initial stages of the program, major emphasis was placed onselecting epoxy resins that are known for their fast curing characteristics.One of these was a specialty resin called Apogen 101 which is supplied bySchaefer Chemicals, Inc. This material tends to produce fast gel and curerates because of the activating effect of methylol groups adjacent to theglycidyl ether functionality of the bisphenol A structure. It was hoped thata small increase in water sensitivity due to the methylol group would be in-consequential for the present application in a highly cross-linked structure.

Moreover, the presence of the methylol groups on the resin might improvewetting and adhesion to the aggregates. A major drawback of this resin isits very high viscosity (quoted as 7,000 to 11,000 cps at 500 C), which makes

major dilution with other resins and diluents mandatory.

Apogen 141 is another member of this family of resins. This is amodified lower viscosity resin with somewhat diminished, but still fast, curerate. Its viscosity, however, is still substantial at 16,000 to 19,000 cpsat 250 C.

~8

A resin of known fast reactivity and of low viscosity (900 to 1500cps) * is Shell Chemical's former Epon 812. This triglycidyl ether of gly-cerine produces highly crosslinked structures and fast reaction rates. Itsproduction has been discontinued, reportedly because of the presence of freeepichlorohydrin, which is toxic. However, some initial experiments werefound to be instructive.

Similar to Epon 812 structurally, but of lower molecular weight andlower functionality as well as without the same toxic hazard, are threeepoxy-functional diluents, including Dow's D.E.R. 732 (55 to 100 cps) andD.E.R. 736 (300 to 60 cps) based on polyglycols, and Ciba Geigy's AralditeRD-2 (15 to 25 cps) based on 1,4-butanediol diglycidyl ether. Another ali-phatic diluent is Epoxide No. 7 (5 to 15 cps) based on predominantly C8 and C10

alkyl groups. This diluent exhibits low volatility and a very low order oftoxicity and is manufactured by Procter and Gamble and marketed byCiba-Geigy.

Two aromatic diluents were also included in the investigation. Oneof these is the diglycidyl ether of resorcinol (300 to 500 cps), exhibiting ahigh degree of reactivity and marketed by Wilmington Chemical Corporation asHeloxy 69 (formerly Ciba Geigy's Araldite ERE-1359, now discontinued). Theother one is phenyl glycidyl ether (10 cps maximum), a monofunctional productof Shell Chemical Company.

A major effort in this program involved standard bisphenol A typeepoxy resins such as Epon 828 and Araldite 509. The former is Shell's liquidepoxy resin of 11,000 to 15,000 cps viscosity. The latter is a compoundedresin coupling Araldite 6010 (similar to Epon 828) and Epoxide No. 7 diluentwith a viscosity of 500 to 700 cps.

Exceptionally fast reactivity is claimed for two low-viscosity,compounded epoxy resins of unspecified structure offered by Celanese PolymerSpecialties Company. The resins are Epi-Rez 5027 (90 to 140 cps) and Epi-Rez50727 (800 to 1100 cps) and are specifically designed to be cured with ali-phatic amines. They are claimed to have excellent wetting characteristics andgood low temperature curing properties.

A specialty, moderate viscosity resin from Diamond Shamrock's Pro-cess Chemicals Division, Capcure Epoxide WR (5300 cps), is recommended forfast- and low-temperature cures. Special mercaptan-based curing agents areavailable for this resin.

A novalac type epoxy resin, D.E.N. 431 (1100 to 1700 cps at 125F), by Dow Chemical Company was included in the early experimental program.Epoxy novolacs are noted for their higher functionality and great strengthand toughness when properly cured. However, their high viscosity requiressubstantial dilution for the purpose of the present application.

A two-part epoxy resin system with curing agent, Epibond 1217A/Bfrom M and T Chemicals, Inc., was briefly investigated and its fast cure rateat room temperature verified. However, the paste-like viscosity renders itsuse impractical in this application.

* All viscosity data at 250 C unless noted otherwise.

9

Limited, exploratory work was also carried out with two cycloali-

phatic epoxides from Ciba Geigy: Araldite CY-179 (350-450 cps) and AralditeCY-183 (350 to 800 cps). The former is an alicyclic diepoxy carboxylate, andthe latter is glycidylated difunctional hexahydrophthalic acid.

5.1.3 Polyamide Cures

Polyamide resins are widely used for the room temperature cure of

epoxy resins. Most commonly, their structures are based on fatty acid-

polyamine condensates, and the curing reaction with epoxides proceeds mainlyby means of the remaining amine functionality. Their distinguishing featuresin comparison to polyamines are:

(a) Lower toxicity and skin sensitization.(b) Lower exotherms in large castings.(c) Greater lattitude in stoichiometry.

(d) Improved flexibility.

Depending on the specific structure, their water sensitivity is somewhat less

than that of amines, especially before cure. However, since larger propor-tions of polyamides than polyamines are required in relation to the epoxy

functionality, this difference tends to be minimized. Moreover, after cure,

the flexibilizing capacity of the polyamides may actually bring about some-what greater water sensitivity than is the case for the highly crosslinked

amine-cured structures.

Two of the standard polyamide curing agents were used in the earlywork on this program, as shown in Table A-1 in Appendix A. Those are EponV-40 (200 to 600 cps at 750 C) and Versamid 125 (7000 to 9000 cps at 750 C)from Shell Chemical Company and General Mills, Inc., respectively (equivalentproducts also available from other companies). A modified amine adduct from

Schaefer Chemicals, Inc., Apogen 256 (300 to 600 cps), was included in this

work for comparison.

The laboratory studies were conducted with and without the additionof an accelerator (DMP-30, Rohm and Haas' tridimethyl aminomethyl phenol).The use of this accelerator was clearly beneficial in formulations containing

the polyamides. Without DMP-30, the amine adduct gave generally faster curerates than the polyamides. However, the gel times of all of these composi-tions were too slow for the present application in spite of the fact that

highly reactive epoxy resins were being used.

This trend continued to hold true for similar compositions shown inTable A-2, but employing D.E.R. 732 and 736 as diluents to reduce the highviscosity of Apogen 101. The three experiments at the top of that table em-

ployed Capcure 38 (450 cps) from Diamond Shamrock, an amido-amine type curing

agent which improves wettability and imparts good adhesion to fresh or oldconcrete. Since this curing agent could not be conveniently compounded with

the Apogen 101 at room temperature, mixing was accomplished after heatingeach of the components to 170 F. Under these conditions, very fast curerates were obtained. However, these findings could not be directly trans-lated to a practical system for use In the field without considerable re-formulating to obtain a sprayable product viscosity under ambient conditions.

10

The experiments described in Table A-3 are based on another re-active, high viscosity resin, the epoxy novolac D.E.N. 431. The same type ofcuring agents discussed above were again investigated, with some of the roomtemperature cures coming within the desired 1/2 hour tack-free time (seeformulation 10-6 in Table A-3). However, the difference in cure rate betweenthe film and the casting illustrates how much the cure of this composition isdependent on the exotherm of the reaction. It may be anticipated, therefore,that the low temperature cure would be inadequate.

In the lower part of Table A-3, the effect of elevated applicationtemperatures on one of the more reactive compositions is illustrated. Al-though cure times at 500 to 770 C were quite good for the cast samples andhigh hardnesses were achieved after aging, these compositions did not performsatisfactorily in wet and in lower temperature environments.

In Table A-4, the effectiveneess of Capcure 38 in Epon 828 compos-itions were explored. While the cure rates were generally inadequate, theywere of a similar order of magnitude as those achievable with the more re-active specialty resins such as Apogen 101, Epon 812, and D.E.N. 431. Basedon these and similar observations, it was concluded that major improvementsin cure rate will need to be obtained from changes in the curing systemrather than from any advantages that the more reactive epoxy resins canoffer. Consequently, the viscosity and cost penalties associated with thesemore reactive resins would probably not be justifiable.

5.1.4 Amine Cures

The reaction between amines containing active hydrogens andepoxides constitutes one of the major branches of epoxy technology. Thebasic reactions according to May and Tanaka * can be summarized by thefollowing three equations:

0 OHR'NH + CH -CHR - b R'NHCH CHR (1)

2 2 2OH

R' NH + CH -CHR R' NH CHR (2)2 2 2 2

0 9R' 2NCH2HR +CH-CHR -R' NCH CHROCH 2iR (3)

• Clayton A. Hay and Yoshio Tonaka, editors "Epoxy Resins, Chemistry andTechnology", Marcel Dekker Inc., New York, 1973, p. 144.

11

Much evidence has been adduced to show that in the reaction between a primary

or secondary amine and an epoxide, reactions (1) and (2) account for all of

the facts and that, in effect, the sum of glycidyl ether and hydroxyl groupsremains constant. These conclusions appear to hold true under certain cond-itions for polyfunctional amine curing agents such as polyethyleneamines, for

instance.

Other work, on the other hand, has demonstrated that etherification

by means of reaction (3) can play a major part in the network formation ofamine cured epoxy resins, and that etherification is favored by certain aminestructures and especially under conditions where less than stotchiometric

quantities of amine hydrogens are present. Other factors favoring reactionscheme (3) are the presence of phenol or acid and high reaction temperatures.

Also significant for the present investigation is the observationthat hydroxyl groups from either water or alcohol accelerate the reactionbetween amines and glycidyl ethers. Moreover, the accelerating effect of aspecific alcohol structure is proportional to its concentration. Not only domost starting formulations of epoxy resins contain some hydroxyl functional-ity, but some moisture will almost always be present on the aggregates and

additional hydroxyl functionality is constantly being introduced during theamine-epoxide cure in accordance with reactions (1) and (2). This factor

together with the gellation effect and the consequent reduction in heat dis-

sipation account for the exotherm observed with almost all epoxy cures.

Little emphasis was originally put on amine cures, because some of

the most effective curing agents were thought to be too water sensitive.However, as the unsatisfactory performance of polyamides became obvious, ex-

periments with amine cured compositions were initiated, as shoun in TableA-5.

All of the curing agents are aliphatic amines except for D.E.H. 39

(20 cps) which is Dow Chemical Company's technical grade of aminoethylpiper-

azine (AEP, also available from other suppliers). Ciba Geigy's DP 152 was

more recently known as HY 9517 (4000 to 5000 cps) and is a proprietary modi-

fied aliphatic amine which has been discontinued. Diamond Shamrock's Capcure

EH-30 accelerator is 2,4,6-tri(dimethylaminomethyl) phenol and is function-ally equivalent to DMP 30. Furfuryl alcohol (FA) was used to solubilize

hexamethylene diamine and was then also included in other formulations as a

control. It apparently also had a significant accelerating effect. In this

group of curing agents, AEP was most effective, followed by the hexamethylene

diamine/FA combination and by diethylene triamine (DETA). Under wet condi-

of magnitude, as had been speculated.

In Table A-6 experiments are summarized with Ciba Ge'gy's modified

aromatic amines, Hardener KY 2969 and Hardener 850 (20,000-26,000 cps). Each

of these curing agents was further modified by forming an adduct with 5 per-

cent by weight of salicylic acid. The acid adduct is reported to increase

the cure rate of epoxy resins, but it unfortunately also increases the visc-

osity of the amines. The adduct of Hardener 850 was somewhat effective under

ambient conditions, although there remained a large spread between the cure

rate of the film and of the casting.

12

As indicated in the earlier section on epoxy resins, Celanese'sEpi-Rez 5027 and 50727 are quite low in viscosity and are recommended forfast cures and for good adhesion to damp concrete. The manufacturer's re-commendations to achieve the fastest cures was to use his proprietary, ac-celerated aliphatic amine, Epi-Cure 874 (75-175 cps). According to the datasummarized in Table A-7, fast cures were recorded at ambient conditions withcastings of both resins. At the lower temperatures, Epi-Rez 5027 failed toperform satisfactorily, while Epi-Rez 50727 performed better but not fully asrequired. In particular, when the resin and curing temperature was at 50 C,immersion in water at 50 C and dry-curing at -250 C left the samples part-ially soft or tacky.

A number of aliphatic and aromatic amine curing agents are offeredby Pacific Anchor Chemical Corporation. Several of these were investigatedin conjunction with Epon 828 in accordance with the data shown in Table A-8.Little description of the chemical structure of these curing agents is pro-vided by the supplier, but some of the recommended end use applications ap-peared most relevant to the present study. Thus, Ancamine 1767 (aliphatic)is considered suitable for fast-setting adhesives and cold weather patchingcompounds; Ancamine AD (aliphatic 1000 to 1200 cps) is recommended for fast-setting civil engineering applications, since it is claimed to provide goodbonds to concrete under cold and damp conditions; Ancamine LT (aromatic 1200to 1400 cps) is claimed to be effective under water and at temperatures downto -5* C; Ancamine XT (aliphatic; 74 cps) can accelerate other amine curesand is effective at low temperatures; Sur-Wet R (polyamine; 4000 to 6000cps) provides good adhesion to wet surfaces by displacing the water on thesubstrate, but its cure rate is not very fast (a similar product is availablefrom 3M Company).

According to the results given in Table 8, fast cure rates (under10 minutes tack-free time in castings) were obtained at ambient conditionswith Ancamine 1767, mixtures of Ancamine 1767 and AD, Ancamine AD, and Anca-mine XT. Cure rates of 15 to 20 minutes in castings were achieved under am-bient conditions with mixtures of Sur-Wet R and Ancamine 11767 or AD.

At cold dry conditions at -25' C, Ancamine AD performed best fol-lowed by Ancamine 1767. Cure rates of 28 minutes and 38 minutes were obtain-ed, respectively, with these two curing agents at 230 C and wet conditions.However, at 5° C and wet conditions, these compositions had not yet curedeven though they had been mixed at room temperature. While not fulfilling allthe requirements completely, these two curing agents, and especially AncamineAD, showed considerable promise.

5.1.5 Mercaptan Cures

Because of the unavailability in the past of polymercaptans of re-latively low molecular weight and high functionality, combinations of such

resins with epoxy materials always required additional curing agents to formfully reacted networks. Indeed, without any accelerator, the reaction pro-ceeds very slowly. However, in the presence of a tertiary amine catalyst,the epoxy-mercaptan reaction can proceed several times faster and at lowertemperatures than the epoxy-amine reaction.

13

The amine-catalyzed epoxy-mercaptan cure may be considered either a

general base catalysis or a nucleophilic catalysis. The reaction scheme ac-cording to the former proceeds through the initial reaction of the mercaptanwith the amine as in reaction (4) followed by reaction with the epoxide inaccordance with (5):

+R 3 N + R'SH __ R'S- + R3NH (4)

2 A 2

+ OH

A+ RNH > R'SCH HCH2OR + R3N (5b)3 2 2 3

The nucleophilic catalysis postulates an initial reaction between the amineand the epoxide followed by a nucleophilic displacement by the mercaptan inaccordance with (6):

O + 9

R3N + CH2-CHCH OR - R NCH H23 R2 2 3 2 6a

B + R'SH > R'SCH 2CHCH 2OR + R 3N (6b)

The investigation of mercaptan curing agents in the laboratoryphase of this program proceeded primarily with the aid of a Diamond Shamrockmercaptan-terminated liquid polymer, Capcure 3-800 (15,000 cps). In thepresence of an amine catalyst, combinations of this reactant with an epoxyresin cure rapidly even at low temperatures, according to the manufacturer.

In early experiments, still using Epon 812 and Apogen 101, tackfreetimes of 2 minutes and 3.5 minutes were obtained in castings and films re-spectively (see Table A-9). Even at -20PC, curing within 5 and 10 minuteswas obtained. Moreover, it was shown by the experiments in Table A-10 thatcompositions based on Epon 828 could be cured nearly as quickly. Underwater, films of these compositions cured within 12 minutes and developedappreciable hardness on aging.

Based on these results, considerable effort was devoted to the

mercaptan curing system. The next step, illustrated by the experiments inTable A-1i, was to develop formulations of lower viscosity with the aid of

reactive epoxy diluents. The fastest cures were obtained with the resorcinoldiepoxide (Heloxy 69). However, while the viscosities of Epon 828 and ofCapture 3-800 were well matched for mixing, the addlition of Heloxy 69 to theepoxy component brought the viscosity into imbalance.

Efforts to correct this viscosity imbalance initially took twodirections, as shown in Table A-12. One of these was to rciort to diluents

14

such as furfuryl alcohol, p-nonylphenol and benzyl alcohol which do not reactwith the mercaptan resin and which can be used to reduce its viscosity.However, while the fast cure rates were maintained, substantial reductions inhardness resulted.

The other direction for viscosity adjustment shown in Table A-12made use of Diamond Shamrock's formulated epoxy resin Capcure WR (5300 cps)and its low viscosity mercaptan reactant Capcure WR-6 (250 cps). The resultswith this system appeared promising.

An alternative direction for viscosity control was to use Araldite509 epoxy resin which is a combination of Araldite 6010 and Epoxide No. 7diluent (see Table A-13). Some reduction in hardness as compared to Epon 828was noted, and some residual tackiness at low temperature curing conditionswas obtained. Further dilution of Araldite 509 with low molecular weightepoxides reduced the hardness considerably, although the cure rate was sub-stantially maintained.

Still another approach to viscosity control was the use of low-viscosity, polyfunctional mercaptans, as summarized in Table A-14. The cur-ing agents selected for this work were Cincinnati Milacron's mercaptate Q-43ester, pentaerythritol tetrakis (mercaptoproprionate) and mercaptate Q-42ester, pentaerythritol tetrakis (thioglycolate). Both of these are alsoavailable from Evans Chemetics, now a W. R. Grace and Company subsidiary.Fast cures of Epon 828 resin was obtained both with the mercaptate ester byitself or in combination with the mercaptan resin Capcure 3-800. Hardnessappeared satisfactory upon aging. Combinations with various epoxy resins aswell as epoxy resin mixtures were cured with Capcure 3-800 and mercaptateQ-43 ester with generally good results (Table A-15). Also illustrated inthis table is the use of Epodil L (100-150 cps), an inert nonreactive liquidhydrocarbon resin diluent. When used in moderation, this diluent did notsignificantly affect either the cure rate or the hardness.

The resin screening program had indicated the versatility of themercaptan cure with a variety of resin and diluent combinations. This curesystem had also given very promising results at low temperatures, under wetcuring conditions at 230 C and 50 C and under elevated temperature applic-ation conditions at 500 C (see Table A-16). Resin components were furtheradjusted, as shown in this table, to provide low-viscosity mixtures of ap-proximately equal volume of epoxy resin to mercaptate curing components forspray application. Formulation 122-1 of Table A-16 was selected for furtherevaluation with aggregates as described more fully below.

5.1.6 Mercaptan/Acrylic Monomer Cures

As a result of the search for diluents for the mercaptan curedformulations, it was found that unsaturated acrylic monomers reacted quicklywith the mercaptan resins. It was found feasible, therefore, to use poly-functional low viscosity acrylic monomers as diluents for the epoxy resinwhen the mercaptan curing system was employed. Under these conditions a dualcure takes place between the mercaptan resin and the epoxide and between themercaptan resin and the acrylic unsaturation.

15

Table A-17 summarizes the results of experiments using trimethyl-olpropane triacrylate (TMPTA) and tetraethyleneglycol diacrylate (TEGDA) in

systems based on Epon 828 and using Capcure 3-800 with or without mercaptateesters as the curing agents. The tack-free times indicated very fast curerates which appeared unaffected by application conditions of -25 C dry and5' C wet.

Additional variations in the ratios of the curing system were ex-

amined and low temperature tests performed as shown in the top portion ofTable A-18. Viscosity adjustment and weight equalization of the componentswas carried out as shown in the lower part of this table and formulation116-2 was selected for spray application evaluation on aggregates.

5.1.7 Polyamide and Amine/Mercaptan Cures

Attempts were also made to harness the rapid-cure characteristics

of the mercaptan resins to boost the performance of epoxy/amide and epoxy/-

amine systems. These experiments were conducted in case the good wettingproperties of the amines or polyamides should be required, depending on the

results of the bonding studies.

As can be seen from the experiments shown in Table A-19, in the

case of the castings, the fast cure rates obtainable with mercaptans were

maintained even when a substantial fraction of the curing system was sub-stituted with Apogen 256 or Versamid 125. Some reduction of cure rate was

discernible in the case of the films. In several of the experiments usingMercaptate Q-43, the hardness of the resins fell well below the Shore D 80range that may be expected. This probably indicates that too high a ratio of

coreactants to epoxy resins was being used.

Table A-20 describes the continuation of this line of inquiry, us-

ing Capcure 38 amido-amine resin. Some decrease in cure rate was observed asthe Capcure 38 ratio was raised, but the faster rates attributable to themercaptan generally dominated, especially in the case of the castings. Atequal levels, Mercaptate Q-43 ester was more effective than Capcure 3-800 in

conjunction with amido-amine curing agent (compare 94-2 and 79-1).

5.1.8 Lewis Acid Cures

Two Lewis acid or cationic catalysts were introduced into the in-vestigation, and the experiments are described in Tables A-21, A-22, andA-23. The catalysts were DP-116 (now renamed KU 195) from Ciba-Geigy andResicure 30 from Ozark-Mahoning Company, a Pennwalt subsidiary. Both ofthese catalysts gave extremely fast cure rates with Araldite 509. Formula-tions with DP 116 cured in 1 to 1.5 minutes on glass plates at room temper-ature. Shore D hardness was ?8 after 10 minutes and 85 after 15 minutes.There was no further increase in hardness with time.

Resicure 30 was even faster than DP 116. As a consequence of thehigh exotherm of the cure, many of the samples became charred. This was the

case in combinations of Regicure 30 and Araldite 509 or Apogen 141, respect-ively. Epon 828 formulated with 10 phr Reaicure 30 cured within 1 minutewithout charring. Shore D hardness was up to 85 within 15 minutes.

16

Experiments with the above systems were also conducted with fur-furyl alcohol as a reactive diluent. Such formulations are of interest be-cause furfuryl alcohol remains very fluid at -25* C. At 25 phr the cure ratewith either DP 116 or Resicure 30 was unaffected when compared to the unmod-ified formulations. However, the cured formulations were noticeably rubberywhen the diluent was present and remained that way on aging.

Araldite DP 116 is used at relatively high levels (25 parts byweight based on epoxy resins). Immersed in water, the upper layer of thefilm which was in direct contact with the water remained tacky and soft for along time, while the layer near the glass surface had hardened. The effectwas evident in formulations with and without furfuryl alcohol, but was morepronounced in the presence of this water-miscible diluent. Similar resultswere also obtained with experiments carried out at low temperatures.

Resicure 30 is used at a level of 10 phr. Hard, tack-free sampleswere obtained within 1 minute on immersion in water as well as in the freez-er. Additional experiments were then run at 5 phr and 2.5 phr based on theweight of epoxy resin. Full cures were also obtained with 5-phr Resicure 30both in water and at low temperatures. Cure times were only a little slower(1.5 to 2 minutes). But the 2.5 level of Resicure 30 was insufficient toprovide good cures, leaving tacky films after 2 hours under water and in air.

Attempts were also made to use Resicure 30 as an accelerator ofpolyamide or amido-amine resin cures. The results shown at the bottom ofTable A-23 indicate that this was not successful. Very likely the Lewis acidwas deactivated or so strongly complexed by the amine functionality that itwas not available for cure at room temperature.

Resicure 30 was carefully considered as one of the curing systemsfor further evaluation. The final decision against its use was based on thefollowing considerations:

(1) The reactivity is too fast even for spray application,and there may not be enough time for the resin to flowout and fill the interstices between the aggregate tothe desired depth.

(2) The application ratio of the two-component system isin the range of 10:1 to 20:1 epoxy resin to curingagent. This is inconvenient to spray, and the systemis very likely to be unforgiving of variations inspray application that might well occur in thefield.

(3) Although good underwater cures have been obtained,there was some evidence that the cure in the very thinsurface layer next to the water may have been incomplete.While this would not detract from the bulk propertiesof the resin, it was a serious detriment to theadhesion of the resin to the wet aggregate (see ResinBonding to Aggregates, Section 5.3).

17

Work had not originally been planned with cycloaliphatic epoxiesbecause of their greater sensitizing tendency as compared to more conven-tional epoxy resins. However, a number of cycloaliphatic epoxy resins havefound acceptance in commerce for special applications. Moreover, the workdescribed above with Lewis acid catalysts raised the question whether thiscuring system in combination with cycloaliphatic epoxides might be suitable

for the present application.

Cycloaliphatic epoxy resins are mainly attractive because of

their low viscosity. The two products selected for this screening study wereCiba-Geigy's Araldite CY 179 (350-450 cps), an alicyclic diepoxy carboxylate,and Araldite CY 183 (350-800 cps), a cycloaliphatic diglycidyl ester resin.Reactions with Resicure 30, with and without methyl nadic anhydride and withdibutyltin dilaurate, were attempted with widely varying results (see TableA-24). With the exception of 73-2, Araldite CY 183 did not cure in these

trials. In experiment 73-2 a 10-minute quiescent initiation period was fol-lowed by a sudden exotherm which charred the resin as it cured. Araldite CY179 cured very quickly in most of the experiments tried, but the reactionproceeded too quickly and was difficult to control.

5.2 Selection of Coupling Agents for Use

in Polymer Concrete

The strength of a filled polymer, including polymer concrete, de-pends to a considerable extent upon the formation of bonds between the filler(aggregate) and the bulk of the polymer. If this were not so, a filledpolymer would be little more than a polymer containing voids, albeit occupiedvoids. On the other hand, if effective bonding is achieved, the strength ofthe final product is typically greater than that of the polymer alone. If

the polymer does not spontaneously bond to the filler, use of a coupling a-gent can be helpful. An effective coupling agent should bond, either by re-action or other attractive mechanism (e.g., solubility in the polymer) to

both aggregate and polymer. In the case of a polymer concrete, this requires

that the coupling agent have two different functionalities.

Among the most widely used coupling agents for filled (rein-forced) polymer compositions are the silanes. In addition, certain titanatesare used for the same purpose. Both of these classes of compounds react in acomparable way with appropriate substrates (aggregates) to permit bonding ofpolymer since both contain organic and inorganic functionality. For example,the silane coupling agents typically have the structure

R-Si(OR')3

where the R-group will be reactive toward the organic polymer molecules whilethe -OR' groups on hydrolysis at the appropriate pH will permit the formationof Si-O-H bond to the oxide functionality of the aggregate. Thus, withglass-fiber reinforced polymers, the coupler would react with the hydrated

silaceous substrate as follows:

R-Si(OR')3 + H 20 ----- ) R-Si(OH) 3 + 3ROH

/OH

R-Si-(OH) + HO-Si-glass--R-SI0-Si-glass (7)3 OH

18

thus anchoring the silane to the glass and leaving the R group available forreaction with the polymer system. Titanium may be substituted for silicon inthe above representation without materially altering the process except forreaction rates and conditions. Both the silica and the limestone or dolomiteaggregates considered for use in the present study should be reactive towardthese coupling agents.

The R-group must be compatible with (i.e., soluble in) or prefer-ably reactive toward the polymer. Since epoxy resins have been selected foruse in this program, appropriate materials can readily be chosen from amongcommercially available materials. The following coupling agents were sel-ected for evaluation based on the criteria noted above.

5.2.1 Union Carbide Silanes

A-100 2 N-CH2CH2CH2-Si(OCH2CH3)3

A-aminopropyl triethoxysilane

A-187 CH -2 H-CH -0-CH CH CH -Si(OCH2 2 2 2 2 H3)3

A-glycidoxypropyl trimethoxysilane

A-189 HS-CH 2-CH2 -CH 2-Si(OCH3)3

X-mercapto-propyl trimethoxysilane (8)

Each of these compounds contains the requisite silyl ether functionality forreaction with the aggregate and a hydrolytically stable organic function,amino, epoxy, and mercapto, respectively, capable of reacting with the poly-merizing epoxy function. In addition, a single titanium derivative fromKenrich Petrochemicals, Inc., was chosen, namely

Ken-React TTS CH 3-C-0-Ti(O-CO-(CH2)14 --CHCH3)3CH 3 CH 3

Isopropyl triisostearic titanate

This material differs significantly from the silane couplers in two ways:there is only one potential bond forming site for reaction with the aggregate(the single isopropoxy group) while the silanes have three, and there arethree isostearic acid groups to provide compatibility with the polymer rather

than the single one of the silanes. However, this titanate is not reactivetoward the epoxy resin as are the silanes and would thus provide bonding onlyby virtue of compatibility with solubility in the resin.

19

WON-

These coupling agents can be applied in two ways: by pretreattngthe aggregate or by combining the coupling agent with the monomer-curingagent mixture. As a matter of convenience to permit controlled hydrolysis atadjusted pH and thus insure reaction with the aggregate, the former methodseems preferable for the present work. It is also to be expected that ag-

gregate pretreatment would make more effective use of the coupling agents,since only a limited time is available before gellation of the resin for thecoupling agent to migrate and bond to the aggregate surface.

Based largely upon the manufacturer's recommendations, the silanecoupling agents were applied to test blocks from 5 volume percent aqueous

solutions adjusted to pH 4.5-5 with acetic acid and used within 1 hour ofpreparation. The blocks were immersed in the solution for one minute,allowed to drain and air dry, and finally cured by heating for 1 hour at 100C. Couplers A-I100 and A-187 dissolved readily in the dilute acetic acidwhile A-189 dissolved only very slowly and required considerable agitiationto achieve complete solution. A similar procedure was also followed fortreatment of aggregates used in the preparation of polymer concrete. Exper-imental results are presented in the sections on Resin Bonding to Aggregates

and Flexural Strength of Polymer Concrete. Only A-189 was used for treatingaggregate based on the preliminary findings with test blocks and was appliedin the same manner.

The titanate complex is not water soluble. Solutions of it inpetroleum ether (5 volume percent) were used for treating the test blocks forI minute followed by drying and curing in the same way as that by which thesilane treated blocks were processed.

5.3 Resin Bonding to Aggregates

5.3.1 General Procedure

For obtaining property data on candidate resin systems, a rapidtesting technique was desirable to facilitate testing multiple samples atclose to one time period. In addition, a technique which would (1) use aminimum amount of hand-mixed resin to avoid exotherm problems, (2) avoid moldstripping problems (adhesion), and (3) provide meaningful engineering data

was desirable. To meet these objectives, a tensile testing technique modeledafter that described in ASTM C321 (Bond Strength of Chemically ResistantMortars) was devised. Figure I illustrates the adhesively bonded cross-blocked tensile set-up which was used.

In preparing the sample, masking tape was applied to the ends ofeach block to keep resin off the loading areas. After the resin was hand-mixed and applied, the specimens were placed in pin alignment fixture to holdthem in proper alignment while the epoxy adhesive cured. The samples werethen placed on the lower U-shaped fixture, the upper fixture was placed abovethe sample, and the assembly was loaded in an Instron testing machine at aloading rate of 0.1 in/m in. To reduce the number of specimens required,

specimens which failed at the bond line were rotated 90 degrees to exposeanother surface for evaluating another rezin cr for testing at another tem-perature. Using this technique, four tests could be conducted with most setsof test blocks.

20

Figure 1. Adhesively Bonded Crossed-Block Tensile Test

Specimen in Testing Rig (Each Block lxlx2 Inches)

21

5.3.2 Substrate Selection

Substrate test block materials chemically similar to the concreteaggregates were used for these screening tests. Silica blocks were cut *from North American Refractories "Narsil" silica bricks which are made fromthe same round silica aggregate deposit (Sharon Conglomerate) used in theconcrete evaluations. These bricks are about 95.5 Si02, 2.6 CaO, 1.0 Fe203,and 0.25 A1203 , have a flexural strength of about 1000 psi, a bulk density of116 to 121 pcf (actual block data), and an apparent porosity of 16.5 to 20.5percent (actual block data equivalent to an absorption of 8.5 to 11.0 per-cent. Due to the heterogeneous nature of the locally available dolomite ag-gregate (fossil deposits), a building stone grade limestone (97.4 CaC03, 1.2MgCO 3) with a fine uniform texture was selected to minimize porosity varia-tions. All limestone test blocks were cut from the same piece of material,which had a bulk density of 141 to 146 pcf (actual block data) and an appar-ent porosity of 15.9 to 16.4 percent (actual block data equivalent to an ab-sorption of 7.0 percent).

5.3.3 Results of Bonding Stud,

An initial test was conducted using an Epon 828 resin formulationdiluted with acrylic monomer and cured with mercaptan resins (71-3, TableA-18). Adhesive bonds were tested "30 minutes after application at roomtemperature. Dry untreated blocks broke in the range of 275 to 360 psi(quartz) and 235 to 270 psi (limestone), which is below the adhesive strengthof the resin bond to the surface. These results verified our expectationsthat epoxy resins would give good bond strengths to dry aggregates.

Adhesion to wet test specimens was of much poorer quality, as mightbe expected. A somewhat wider range of formulations was evaluated on wetsilica and limestone blocks in order to discover possible differences intheir adhesion performance to wet aggregates. As can be seen from Table 1,none of the formulations were satisfactory. The resin cured with the Lewisacid catalyst, however, had exceptionally poor adhesion, as had been antici-pated (see Resin Formulation and Screening, Lewis Acid Cures, Section 5.1.8).

Three silane and one titanate coupling agents were used to treatsilica and limestone blocks in accordance with procedures described earlier,and two epoxy/mercaptan formulations, one with and one without trifunctionalacrylate monomer, were employed to evaluate adhesion to treated, wet silicaand limestone surfaces. From the data in Table 2, it can be seen that sign-ificant improvement in bond strength is achievable with several of thecoupling agents. The best results were obtained with the mercaptate-functional silane (A-187) and the epoxy-functional silane (A-189). Whenepoxy composition 71-3 was used on wet substrates pretreated with A-187, thebond strength was greater than the breaking strength of the blocks. A sign-ificant improvement in bond strength was also obtained with the titanate whenFormulation 58-1 was used, but not in the tests employing Formulation 71-3.The aminofunctional silane was ineffective in all instances.

To avoid contamination all test block materials were cut with a diamond

saw and then surface ground flat and paralleled using only tap water asa lubricant.

22

U

0 4-

0 ~ 0co Ch.

z 000

w 01

102

00 In A A

C) 4) "00 3. ).4 0 m mC00 v..4 Z 4 41%

z Ua 000 -Ht o Q -U

0 000w

U7 41 d)

41

&JJ t1

It4 14'0 cal u.

I- 0I"

M0

.741

0

- 41 r4 -

4) 0 t -4

G)

0 S0o . to . .414' .0

'Uv

II..

23

TABLE 2. ADHESIVE BOND STRENGTHS (PSI) OF EPOXY COMPOSITIONS TOTREATED AND UNTREATED SILICA AND LIMESTONE SUBSTRATESUNDER WET CONDITIONS AT 730 F

Epoxy Formulation 5 8-1 (b)

Silica Blocks Limestone Blocks

Untreated, wet, 730 F 69 22

Treated(a), wet, 730 F

Silane A-11O0 35 17Silane A-187 112 125Silane A-189 264 150Ken React TTS 64 191

Epoxy Formulation 71-3(c)

Untreated, dry > 316 > 247

Untreated, wet, 730 F 16 4

Treated(a), wet, 730 F

Silane A-1100 13 31Silane A-187 > 242 > 117Silane A-189 242 101Ken React TTS 15 30

(a) Silane A-1100 = Union Carbide's gamma-aminopropyltriethoxysilane

Silane A-187 = Union Carbide's gamna-glycidoxypropyltrimethoxy-silane

Silane A-189 - Union Carbide's gauma-mercaptopropyltrimethoxy-

silane

Ken React TTS - Kenrich Petrochemicals' isopropyltriisostearic

titanate.

(b) Bisphenol-A/epichlorophydrin diluted with 1,4-butanedioldiglycidyl ether and rured with mercaptan-functi+onal resin(see Table A-li).

(c) Combination of bisphenol-A/epichlorohydrin epoxy resin and tri-functional acrylate monomer cured with mercaptan-functionalresin (see Table A-IF).

24

Similar tests were also conducted with the wet substrates condi-tioned at 400 to 450 F. The data shown in Table 3 lead to very similar con-clusions as those reached with the tests conducted at room temperature. Theepoxide- and mercaptan-functional silanes were again among the most effectivecoupling agents. Also consistent with the previous findings, the titanatecoupling agent was effective only with formulation 58-1 and only on limestoneblocks. Variations between the two sets of data do not appear to follow a

discernible pattern and are probably attributable to data scattering perhapsdue to nonuniformity of resin application.

5.4 Adhesive Strength to Asphalt and Concrete

5.4.1 Substrate Selection

Adhesive strength of the resin to asphalt and Portland cement con-crete was evaluated using the tensile bond strength test procedure described

in the preceding section for screening resins, except that asphalt and Port-land cement concrete test block materials were used. The asphalt concretetest material was a paving block manufactured by Hastings Pavement Company,Lake Success, New York, and used for industrial flooring. It has an asphaltcontent in the 5-7 weight percent range of normal asphalt road mixes andcontained limestone aggregate graded to provide a void content of about 0.5

percent (not verified) compared to 3 to 9 percent in typical road mixes. Theasphalt used in these blocks had a higher softening point than most roadmixes, having less than 30 mm penetration at room temperature (ASTM D-946)compared to 70 to 80 mm for road mixes. The Portland cement concrete testmaterial was a patio block mix containing a local dolomite aggregate. Sinceit was obtained from a local building supply yard, its exact source and pro-perties were unknown.

5.4.2 Results of Adhesion Tests

The adhesion evaluation was carried out with formulation 71-3 whichis based on Epon 828 diluted with trimethylolpropane triacrylate and curedwith mercaptan resins (see Table A-18). The results are summarized in Table4.

Under dry application conditions, both at 730 F and at 140 F, thetest blocks broke before the resin/substrate bond gave way. It was encour-aging to find that a thin resin layer could be cured in 1/2 hour at 140 F toan adhesive and cohesive strength greater than the breaking strength ofeither asphalt or Portland cement concrete.

Under wet application conditions adhesive failure occurred in allof the samples, although the strength developed within 1/2 hour was abouttwice the minimum strength of 40 psi which had been specified. A furtherimprovement of about 50 percent of the adhesive strength under wet conditionscould be obtained by incorporating 0.2 percent of a mercaptan-functionalsilane coupling agent into the resin mixture. Such a coupling agent could beadded to the mercaptan curing component and may be expected to have goodstability provided water was excluded. Alternatively, an epoxy-functionalsilane coupling agent could be incorporated in the epoxy resin componentunder exclusion of moisture.

25

TABLE 3. ADHESIVE BOND STRENGTHS (PSI) OF EPOXY COMPOSITIONS TOTREATED AND UNTREATED SILICA AND LIMESTONE SUBSTRATESUNDER WET CONDITIONS AT 40 TO 450 F

Epoxy Formulation 5 8-1(b)

Silica Blocks Limestone Blocks

Untreated, wet, 73' F 69 22

Treated(a), wet, 40-45* F

Silane A-1100 87 90Silane A-187 > 338 120Silane A-189 172 192Ken React TTS 55 193

toxy Formulation 71..3(c)

Untreated, dry, 730 F > 316 > 247

Untreated, wet, 73* F 16 4

Treated(a), wet, 40-450 F

Silane A-1100 38 30Silane A-187 90 98Silane A-189 130 113Ken React TTS 15 30

(a) Silane A-1100 = Union Carbide's gauna-aminopropyltriethoxysilane

Silane A-187 Union Carbide's gamma-glycidoxypropyltrimethoxy-silane

Silane A-189 - Union Carbide's gamma-mercaptopropyltrimethoxy-silane

Ken React TTS =Kenrich Petrochemicals' isopropyltriisostearictitanate.

(b) Bisphenol-A/epichlorophydrin diluted with 1,4-butanedioldiglycidyl ether and cured with mercaptan-functlonal resin(see Table A-I1);

(c) Combination of bisphenol-A/epichlorophydrin epoxy resin and tri-functional acrylate monomer cured with mercaptan-functionalresin (see Table A-18).

26

TABLE 4. ADHESIVE BOND STRENGTH (PSI) OF EPOXY COMPOSITION71-3 TO ASPHALT AND PORTLAND CEMENT CONCRETE

Adhesive Strength(a)

Asphalt ConcreteFailure Failure

Application Conditions psi Mode psi Mode

Dry 730 F 204 Block 365 Blockfailure failure

Dry 140 F 204 Block and 362 Blockcohesive failurefailure

Wet 730 F 83 Adhesive 80 Adhesivefailure failure

Wet + 730 F 115 Adhesive 135 CohesiveA-189(b) failure failure

(a) Tested on the Instron in a tensile mode at 0.1 in/mmn crossheadspeed 30 minutes after application.

(b) 0.2 percent by weight added to the resin.

27

5.5 Flexural Strength of Polymer Concrete

5.5.1. Aggregate Selection/Characterization

Round quartz pebbles and crushed dolomite (CaCO3 MgCO3 mineral)aggregates obtained from local sources (Parry Company, Chillicothe, Ohio, andMarble Cliff Quarries, Columbus, Ohio, respectively) were selected for thisprogram as being representative of two types of natural aggregates with dif-ferent physical and chemical characteristics. Three size ranges (3/4-inch x1/2-inch, 1/2-inch x 4-mesh, and 4- x 8-mesh) were prepared, but based onpermeability considerations, only the coarsest one (3/4-inch x 1/2-inch) wasused in actual specimen fabrication work.

Permeability coefficient of the aggregate blend used was believedto be the key factor in determining how rapidly the epoxy resin would pene-trate the voids. Since epoxy formulations with viscosities in the 103 cprange were being developed which gelled in about 2 minutes, a target penetr-ation rate (flux) of 4 inches/minute was assumed necessary for preparation ofan 8-inch-thick slab. For the special case of free downward flow of a liquidcolumn, * Darcy's law relating permeability coefficient and fluid flow para-meters becomes:

k - permeability coefficient (darcys) = (1)Pg

where 10- 7 N sec- viscosity (cp 2 )

cm

v flux (cm/sec)

p - fluid density (g/cc)

g - 9.8 m/sec2

Substituting a viscosity of 10 cp, a flux of 4 inches/minutes, and a Oensityof 1 g/cc yields a required permeability coefficient of about 1.7 x 10'darcys. Permeability coefficient data reported by Bo, et al** for varioussited aggregates have been plotted in Figure 2 and extrapolated to the rangeof interest to aid in selecting an aggregate size with an appropriate perme-ability coefficient. Based on the above calculation, an aggregate size ofabout I inch is required to obtain penetration rates compatible with resincharacteristics. Preliminary epoxy penetration experiments indicated the 3/4x 1/2-inch aggregate size could be readily penetrated by the resin, but thatthe 1/2 x 4 mesh size caused puddling and air entrapment. This observationis consistent with the dramatic change in permeability coefficient with sizeshown in Figure 2.

The 3/4- x 1/2-inch aggregates selected had void contents of 35 to36 percent (quartz) and 42 percent (dolomite) after consolidation by tablevibration (Table 1). Minor reductions in void volume (about 2 percent) wereobtained by blending this size with the 1/2- x 4-mesh fraction, but this

Recomended Practice for Determining Permeability of Porous Media,American Petroleum Institute, RP 27, 3rd edition, August 1956.

** Bo, M K., Freshwater, D. C., and Scarlett, B., "The Effect of ParticleSize Distribution on the Permeability of Filter Cakes", Trans. lIstChem Engrs., 43, T228 (1965). 28

lo0 5

3 /4-inchI5/8-inch10~ j 1/2-inchI 3/8-inch

14I ye

20 3 6 esh

102 Experimental data from Bo, et al.

100.1 1.0 10 100

Particle Diameter, on

Figure 2. Effect of Aggregate Diameter on Permeability

29

amount was not considered beneficial enough to compensate for penetrationpenalties. Similar improvements should be obtainable through the use ofslightly coarser aggregates (1-1/2 inches x 3/4 inch), but significant re-ductions in the void volume (e.g., 20 percent range) would require a 1:10aggregate size ratio. * The use of aggregates in the 5-inch size range (toget minimal voids) was beyond the scope of this program but appears to be afeasible method of minimizing resin requirements in the field.

5.5.2 Specimen Preparation Procedure

Beam specimens 4x4x14 inches were used to evaluate theresin-concrete system. The beam molds were lined with polyethylene sheet andthe 3/4x1/2-inch aggregates were vibratory compacted about 10 seconds to ob-tain a low void volume (see Table 5). The molds were then conditioned at thedesired test temperature, removed from the conditioning chamber, and sprayedwith a stream of resin applied to one edge of the beam to allow venting ofair from the other edge. fixing and application of the two-component resinsystem was accomplished with the aid of a Graco Bulldog Hydra Cat airlessspray unit. After stripping the sample from the molds, the sprayed edge wasplaced down in the test fixture such that it was loaded in tension. Figure 3shows a sample being prepared, and Figure 4 shows a sample being tested inthird point loading. **

5.5.3 Results of the Flexural Strength Tests

Both of the two resin systems selected for this evaluation werebased on Epon 828 resin and cured with a mixture of mercaptate resins(Capcure 3-800 and Mercaptate Ester Q-43 for viscosity control of the curingresin component). The two resin systems differed primarily in the diluentused for the epoxy resin component: (a) a difunctional resorcinol diglycidylether, in accordance with formulation 58-IM (Table I and Table A-1l), and (b)a trifunctional acrylate, in accordance with formulation 71-3 (Table A-18).Some additional formulation adjustments were made to equalize the volumes ofthe epoxy resin and the curing resin components more closely. The finalformulations used in the flexural strength evaluations were 122-1 and 116-2(in Tables A-16 and A-18, respectively). TheIr compositions, by weight, wereas follows:

122-1 116-2Part I

Epon 828 33.3 35Trimethylol propane -- 15

triacrylateHeloxy 69 16.7 --

Part II

Capcure 3-800 27.2 22.5Mercaptate Q-43 18.1 22.5Capcure EH-30 4.1 ,_1

100.0 i00.0

Powers, T. C., "Geometric Properties of Particles and Aggregates",

PCA Bulletin 174, pp. 4-17 (1964).

*k ASTM C78 except that the inner span length was zeduced to 3 inches tofacilitate uniform loading at the outer spva, strength calculations

3Pa4were made by the general equ.:t.on a for tese beams

30

TABLE 5. PROPERTIES OF AGGREGATES EVALUATED

Unit Bulk Specific Void ApparentPounds(a) Weight, Gravity(c) Voluye, Sp. Gravity, Absor tfon,

Aggregate Size Compaction Gross Net pertb) g/cc pcf 0/0b

glcc(c) o/oT)

Parry Quartz 3/4 . 1/2 (c) loose(d) 29.50 23.90 97.15 2.61 162.9 40.4 2.63 0.28(round)

Parry Quartz 1/2 x 4 (M) loose(d) 29.50 23.90 97.15 2.61 162.9 40.4 2.63 0.28(round)

Parry Quartz 60 C, 40 M loose(d) 30.20 24.60 100.0 2.61 162.9 38.6 2.63 0.28(round)

Parry Quartz 3/4 x 1/2 (c) loose(e) 29.2 23.6 95.9 2.61 162.9 41.1 2.63 0.28

(round)

Parry Quartz 1/2 x 4 (M) loose(e) 29.4 23.8 96.7 2.61 162.9 40.6 2.63 0.28

(round)

Parry Quartz 3/4 x 1/2 (c) vibrated(d) 31.25 25.65 104.3 2.61 162.9 36.0 2.63 0.28(round)

Parry Quartz 1/2 x 4 (M) vibrated(d) 31.55 25.95 105.5 2.61 162.9 35.2 2.63 0.28(round)

Parry Quartz 60 C. 40 M vibrated(d) 32.40 26.8 108.9 2.61 162.9 33.1 2.63 0.28(round)

Parry Quartz 3/4 x 1/2 (c) vibrated(e )

31.6 26.0 105.7 2.61 162.9 35.1 2.63 0.28(round)

Parry Quartz 1/2 x 4 (M) vibrated(e) 31.8 26.2 106.5 2.61 162.9 34.6 2.63 0.28

(round)

Marble Cliff 3/4 x 1/2(c) loose(e) 25.0 19.4 '8.9 2.48 154.7 49.0 2.71 3.3

Dolomite(crushed)

Marble Cliff 1/2 x 4 (M) loose(e) 25.4 19.8 80.5 2.48 154.7 48.0 2.71 3.3Dolomite(crushed)

Marble Cliff 60 C, 40 M loose(e) 26.4 20.8 84.6 2.48 154.7 45.3 2.71 3.3

Dolomite(crushed)

Marble Cliff 3/4 x 1/2(c) vibrated(e) 27.7 22.1 89.8 2.48 154.7 41.9 2.71 3.3

Dolomite(crushed)

Marble Cliff 1/2 x 4 (M) vibrated(e) 28.2 22.6 91.9 2.48 154.7 40.6 2.71 3.3

Dolomite(crushed)

Marble Cliff 60 C, 40 M vibrated(e) 29.1 23.5 95.5 2.48 154.7 38.3 2.71 3.3Dolomite(crushed)

(a) Weight of container having a factor of 4.065/ft3 per ASTM C29

(b) ASTM C29 except for compaction procedure

(c) ASTM C127

(d) Operator A (10 seconds vibration)

(a) operator 3 (15 seconds vibration)

31

Figure 3. Application of Resin With Airless SprayEquipment on Aggregate-Loaded Test Trays

Figure 4. Flexural Testing of Beams

32

After initial adjustments were made on the Graco spraying equipment, both ofthese formulations could be applied at room temperature without heating. Itwas necessary, however, to insert a static mixer in front of the spray gun toassure that the two components were well mixed. Attempts to spray withoutthe static mixer resulted in inadequate blending of the components as evid-enced by slow cure and poor hardness properties. Filling of the beam moldrequired about 3 to 3-1/2 minutes which is close to the gellation time ofthese resin systems. Thus, the resin that had been sprayed first and had

sunk to the bottom of the mold was about to gel when the filling of the moldwas completed. Timing for testing purposes, however, was started when thefirst resin stream was applied to the aggregates in the molds.

The results of the flexural strength tests are given in Table 6.The initial tests were run on untreated aggregates. All subsequent runs werecarried out with aggregate treated with Union Carbides A-189, marcapto func-tional silane, in accordance with the procedure described in section 5.2 onSelection of Coupling Agents for Use in Polymer Concrete.

The rapid gellation and curing of the resins resulted in a highexotherm which reached its peak from 45 minutes to 1 hour after spraying.Both the 1/2 hour and the 1 hour tests fell into this time region when theresin concrete was at high temperatures and its strength well below theoptimum that could be reached after cool-down. Temperature measurements ob-tained with a surface pyrometer ranged from 1100 to 1400 F within this timeinterval. Internal temperatures were undoubtedly higher.

To illustrate the temperature effect on the flexural strength ofthe resin concrete, measurements were obtained on several test bars 24 hoursafter spraying. The data near the bottom of Table 6 illustrate that flexuralstrengths around 1500 psi can be obtained with dry treated quartz aggregateand around 1200 psi with similar wet aggregate. Moreover, when similar test

bars were quenched in ice water starting 45 minutes after resin application(see Runs 38 and 39), the flexural strength again fell near 1500 psi.

It is concluded from this that the low flexural data are attribut-able to the high exotherm and the heat retention of the test bars rather thanto incomplete cure. This is further supported by the observation that testbars prepared with aggregate at 400 F and others at 100 F had higher flexuralstrengths than comparable samples prepared at room temperature. Clearly, thegreater heat sink provided by the cold aggregates lowered the temperature ofthe bars at the time of testing and thus resulted in higher flexural strengthmeasurements. In fact, the flexural strength measurements at the 100 F cond-itions might have been higher, since penetration of the resin into the coldaggregates was incomplete, that is, the lower regions of the test bars wereincompletely covered with resin.

Other observations which emerge from the data in Table 6 include:

(1) Flexural strengths of resin concrete based on quartz arehigher than those of bars prepared with dolomite (in all but one instance,where the difference is essentially negligible).

33

TABLE 6. RESULTS OF FI.XURAL STRENGTH MEASUREMENTS

Application Testing Flexural

Test Application Temperature. Time, Strength., gregate Conditions F hr pat

Formulation 116-2

Dolomite Dry RT 1 1462() Dolomite Dry RT 1 1763 Quartz Dry RT 1 264

6 DolomiLte Dry IT 1 281

7 Dolomite Dry RT 1 285

5 Quartz Dry RT 1 272

Formulation 122-1

8 Dolomite Dry RT 1 184

9 Dolomite Dry Ri, 1 206

10 Quartz Dry IT 1 352

11 Quartz Dry RT 1 387

12 Dolomite Dry 10 1 40613 Dolomite Dry 10 1 364

14 Quartz Dry 10 1 621

15 Quartz Dry 10 1 590

16 Dolomite wet RT 1 3518 Dolomite wet RT 1 45

17 Quartz Wet IT 1 230

19 Quartz wet IT 1 238

30 Dolomite Dry RT 1/2 148

31 Dolomite Dry RT 1/2 13828 Quartz Dry IT 1/2 18729 Quartz Dry RT 1/2 194

34 Dolomite Wet RT 1/2 29

35 Dolomite wet IT 1/2 38

32 Quartz Vet IT 1/2 354

33 Quartz wet IT 1/2 220

24 Quartz Dry I! 1 309

25 Quartz Dry RT 1 429

26 Quartz Dry 40 1 626

27 Quartz Dry 40 1 468

36 Quartz wet 40 1 354

37 Quartz wet 40 1 220

38 Quartz Dry RT 1 (b) 1568

39 Quartz Dry RT 1 (b) 1438

20 Quartz Dry RT 24 1638

21 Quartz Dry RT 24 1298

22 Quartz Wet RT 24 1147

23 Quartz wet RT 24 1263

(a) Untreated aggrc&nte ; all other agrreates treated with mercaptoftinctional

sllame, A-169.

(b) Samples wL.re quenched In ice water ojr 15 mlnutes beforr, testing.

34

(2) Flexural strengths obtained with wet quartz are only moder-ately lower than comparable data obtained with dry quartz. In contrast,flexural strengths decline drastically between wet and dry dolomite.

The hypothesis that can best support these observations is that the couplingagent treatment was successful with the quartz aggregates, but not with thedolomite. However, it was established during the bonding studies of resin tosilica and limestone blocks that silane A-189 was beneficial for both sur-faces. It would appear to be necessary, therefore, that the coupling agenttreatment be optimized separately for each substrate.

5.6 Durability

5.6.1 Background

A method of evaluating the relative resistance of the resin con-crete to rutting raveling and/or wear from aircraft wheel traffic was re-quired in this program. ASTM methods for abrasion testing of concrete con-sidered for this purpose included a sand blast test (C 418), a rotatingabrasive disc (C799, procedure A), a rotating cutter wheel (C779, procedureBO), and a rotating ball-bearing test (C779, procedure C). However, none ofthese test facilities were available at the contractor facilities or at thePortland Cement Association Laboratories. Contacts with the Army Corps ofEngineers Waterways Experiment Station revealed two somewhat simpler abrasiontest methods (a rotating slurry of steel balls and a single rotating cutterwheel), neither of which imposed high shear forces associated with landing orbraking aircraft. A rotating abrasive cup wheel (a test used previously bythe contractor in comparing the wear resistance of various orthotropic bridgedeck pavements) was considered but was rejected on the basis that the con-tinuous contact area of the wheel would not provide localized impact forcesto the coarse textured resin concrete mix. A rolling grinding wheel techni-que was considered attractive in that it would more closely duplicate thetype of stress actually imposed. However, this technique would have requireda rather elaborate experimental set-up to maintain a constant wheel load us-ing wheels of reasonable size (1-inch x 6-inch diameter), and miniature rot-ating wheel test equipment available at the contractor facility (TaberAbrasive Tester) was not compatible with the 3/4-inch x 1/2-inchaggregate in the concrete.

5.6.2 Test Procedure

Due to the lack of available standard test equipment, and/or con-cern for the appropriateness of these techniques, a rotating rod abrasiontest procedure was devised for this program. The concept provides for re-latively large contact area of a sliding abrasive material (similar to ASTC779, procedure A), but utilizes simpler equipment and imposes Impact forces(similar to ASTM C779, procedure C). Figure 5 illustrates the device, whichis simply three 1-inch diameter hardened tool steel rods welded to the peri-phery of a threaded adapter fitting a commercial diamond core drill motor.The rods form a circular tract about 3-1/2 inches outside diameter and 1-inchwide, permitting tests to be run on portions of 4x4x14-inch beam specimens.The drill motor rotates at 500 RPM (no load) and provides a total dead weightof 47 pounds.

35

Figure 5. Device for Durability Tests

Figure 6. Asphalt Concrete Sample

After Durability Testing

36

Initial experiments with the rods revealed that little (<0.1 inch)or no wear but considerable heat build-up occurred in a 30-minute test periodwhen the rods were run dry and without abrasive material feed. To acceleratethe test and keep the specimens from heating up, a 7-inch-diameter bottomlessplastic beaker about 7 inches high was cemented to the specimen with siliconerubber to contain a 1/2-inch layer (dry) of ASTH C109 sand and sufficientwater to give a one-inch deep slurry. With the wet sand slurry, significantwear data could be obtained in 5-minute test periods, at which time the usedsand was removed and a fresh sand slurry introduced. Figure 5 illustratesthe entire test rig, showing the core drill apparatus and the plastic beakerin position. Figure 6 shows an asphalt concrete specimen (with beaker re-moved) after being tested several times.

5.6.3 Control Materials

In addition to the epoxy mixes, several control materials weretested to provide comparative data. The asphalt paving brick and Portlandcement patio block materials used for adhesive tests were also evaluated fordurability. (These materials are described in a previous adhesive test sec-tion of this report.) In addition, a Portland cement concrete test beam madefrom the same strength of Marble Cliff dolomite aggregate (w/c - 0.45 and28-day flexural of 675 psi) was also evaluated. The epoxy concrete specimenswere about one week old at the time the abrasion tests were conducted. Allof the materials evaluated wore smoothly, and no evidence of aggregate pullout was noted.

5.6.4 Results

Figure 7 is a plot of the abrasive wear results obtained from thevarious materials tested. A total of 6 tests were run on 4 epoxy concretetest specimens, but the wear rates of the 3 quartz and 3 dolomite aggregatesamples were essentially so similar that the data is simply shown as a bandrepresenting the widest data spread. Compared to either the asphalt block ora conventional Portland cement concrete mix (both with dolomite aggregate),the epoxy concrete mixes wear only about one-fourth as fast. The pooragreement between the two asphalt specimens tested may be due to the edges ofthe steel rods being flatter, and possibly sharper, during the first fewtests of the series, * but the two concrete patio slabs were the second andsecond last specimens tested, yet gave excellent agreement. Material vari-ability is another possible cause for this discrepancy.

Although only a limited number of tests were conducted with eachmaterial, the results clearly show that the wear resistance of epoxy concretemade from the preferred epoxy formulation (122-1) is at least twice that ofasphalt or Portland cement concrete control materials after being fully cur-ed. Time constraints imposed near the end of the program precluded abrasiontesting beams only 30 to 60 minutes old, but somewhat similar data might beexpected since the water slurry would tend to cool the beam surface somewhat.Although the relative performance of the epoxy concretes appears promising,

*The leading edges were rounded back about 1/16-inch at the conclusionof the test sequence.

37

II

1.5/ 0 Portland Cement ConcretePatio Slabs (Dolomite)

13 Hastings Asphalt Block(First specimen tested)

0 Portland Cement Concrete w/c 0.45- Hastings Asphalt Block

(Lost sample tested)h Data Bond for Epoxy Concrete Specimens

8,9,10 and II (Epoxy 122-I, 2 dolomite,2 quartz)

I

0.5

0 S30 45Time, minutes 0

Figure 7. Abrasive Wear of Concrete Materials Versus Time

3 f

full-scale rolling wheel field tests would be required to properly assess thedurability of any material ultimately developed.

5.7 Storage Stability of Resin Systems

Epoxy resin components and mercaptan components corresponding tothe several formulations of interest were monitored for several weeks withrespect to viscosity changes under various environmental conditions. Resinswere stored at room temperature, at -20 C, and at 600 C. The latter condi-tion will be rarely, if ever, encountered in the field. It represents anaccelerated aging test and may indicate the storage stability over a longertime period at more moderate temperatures. Such extrapolations, however, arealways risky, since reactions may be initiated at the higher temperatureswhich, under more modest temperature conditions, may not occur. The accel-erated aging data can be used, however, as an alert to possible storage pro-blems and to indicate the need for more extensive storage stabilityevaluations.

Resin mixtures were stored in glass jars. No effort was made toexclude oxygen from the head space. Periodically, the containers werebrought back to room temperature, and their viscosity was determined with aBrookfield viscometer. The high sensitivity of these measurements to smalltemperature fluctuations was not at first recognized, which accounts for thescattering of the data in some of the experiments. Overall trends can stillbe recognized, however.

The epoxy formulations in Table 7 appear to be stable under theaccelerated aging conditions. No trend of viscosity increase can be detectedwith the Epon 828/acrylic monomer mixture. There may be a very slow visco-sity rise in the case of the Epon 828/Heloxy 69 mixture. However, using therule of thumb that every 10 C in reaction temperature approximately doublesthe rate of reaction, the viscosity change in this resin mixture does notappear alarming and a useful resin life in excess of 1 year appears likely.

The observed increase in viscosity in the mercaptan components isof real concern. It is especially high in formulations containing MercaptateQ-43 Ester but is also significant in the case of the Capcure 3-800/EH-30mixture. There is no indication that Capcure 3-800 has a storage stabilityproblem, and it must be concluded that the presence of the tertiary amine,Capcure EH-30, creates the problem. *

In the relatively short temperature storage tests no clearly de-finable trend of viscosity increase could be identified, and it remains to beestablished whether the instability found at 600 C corresponds to a shelflife stability problem under more moderate temperatures. If this is thecase, an alternative accelerator may need to be identified or a means tostabilize the mixture may need to be devised.

* Recent information received from the manufacturer claims good shelf life

stability of Capcure 3-800/EH-30 mixtures up to 50* C.

39

TABLE 7. ACCELERATED AGING TEST OF RESIN COMPONENTSAT 600-C

Reference Table A-16 A-18Formulation 116-1 116-2

Part I Part II Part I Part II

Epon 828 66.7 70

Heloxy 69 33.3

Trimethylolpropane 30triacrylate

Capcure 3-800 90 43.5

Mercaptate Q43 43.5

Capcure EH-30 10 13

Day of MeasurementTest Temperature, F Brookfield Viscosity Data, cps

0 -- 2870 18,840 1896 3,470

7 -- 3040 19,760 1919 6,630

14 73 2370 14,200 1552 8,930

21 72 3200 24,600 2008 --

28 71 4100 28,720 2168 18,080

35 72 3880 29,200 2140 20,400

42 72 4080 29,520 2072 23,000

49 76 3140 20,280 1388 18,720

56 73 3960 34,880 2024 28,320

40

SECTION VI

EQUIPMENT REQUIREMENTS

Assuming that the repair effort with polymer concrete is directedtowards damage of craters about 20-foot diameter and assuming that up to 10craters may need to be repaired, some general conclusions can be reachedregarding equipment needs. The following analysis illustrates situationsrequiring either a 12-inch or a 6-inch deep repair cap:

Depth of repair cap, inch es 12 6Volume of repair cap, ft 3 314 157Required resin volume, (a) ft3 110 55

gal 822 411

Required resin flow rate, gal/minResin application, 5 minutes 164 82

10 minutes 82 4115 minutes 55 2830 minutes 28 14

(a) Assuming 35 percent void volume.

If airless spray equipment of the type used in this project is to beemployed, custom-made units will most likely be required to handle thenecessary flow rates. The supplier of the equipment used in this study, Graco,Inc., of Minneapolis, Minnesota, has supplied units up to 20 to 30gallons/minute capacity in the past by special request. Cost of such units,without heaters, is estimated in the $8,000 to $20,000 range, but economies areundoubtedly possible in the construction of large numbers of such units.

While the 20 to 30 gallon/minute application unit is the largest thatwas supplied in the past, units of greater capacity can undoubtedly bedesigned. Moreover, it is quite feasible to use two or more units in tandem,either feeding into a single hose or running multiple hoses into as manynozzles and letting the streams blend on the repair cap. Static mixers(similar to the one used in this effort) are recommended in favor of mechanicalmixers because of their better wear and low maintenance characteristics andsubstantially lower costs. Large static mixers are readily available and arenot expected to pose any flow limitations.

In addition to the airless spray units, a power supply must beavailable to provide about 100 psi compressed air to the air pumps. Higherpressures are not necessarily desirable, since the equipment typically workson a 34:1 fluid-to-air ratio (compression ratio). Other supportive equipmentmust include a supply truck with dual tanks of a minimum of 200 'r 400-galloncapacity each, depending on depth of repair cap, to repair one crater, andstationary storage tanks of 2000- or 4000-gallon capacity for each resincomponent to service 10 craters. Provisions must also be made to lay down afluid barrier (sand, nonwoven fabric, plastic film) underneath the gradedaggregates of the repair cap.

41

It is assumed that appropriate quantities of aggregates (38 to 76cubic yards for the 10 craters in the above example) will have been pretreatedwith coupling agent and will be stored outdoors under tarpolines. Resinstorage should be near 700 F to avoid excessive viscosity increases in coldweather and to assure maximum storage life in hot weather. Underground storagemay be advisable.

4

42

SECTION VII

CONCLUSIONS AND RECOMMENDATIONS

This research program has succeeded in identifying two-componentepoxy resin formulations of low viscosity that can be applied in equal ratio ofresin to curing agent components by means of commercially available airlessspray equipment. The formulations selected for full evaluation are based onmercaptan curing systems and can utilize a variety of diluents for viscosity

control.

These resin systems typically set up within 3 to 4 minutes aftermixing at temperatures around 730 F. Good cures can be achieved within 1/2hour of mixing in wet environments down to 50 C and in dry environments down to-25* C. Good adhesion to silica or limestone aggregates is obtained under dryconditions without the aid of coupling agents. Under wet conditions, the useof coupling agents is necessary to obtain good adhesion.

On test blocks in the laboratory, several organofunctional silanecoupling agents worked well on both substrates. A mercaptan functional silanewas selected for further evaluation. However, good adhesion with this couplingagent on spray application under wet conditions was only achieved in the caseof quartz aggregates. In the case of dolomite aggregates results were notsatisfactory, and either a different application method or a different couplingagent should be employed.

Adhesion performance of the resin systems to asphalt and Portlandcement concrete was several times greater than the minimum requirement of 40psi when applied under dry conditions. Under wet application condition, theadhesion was still twice the minimum requirement without the aid of couplingagents. When 0.2 percent of a silane coupling agent was added to the resin, animprovement of approximately 50 percent was obtained in adhesive strength tothe wet substrates.

Durability tests were performed with the aid of a specially devisedprocedure which had been approved by the project monitor. The performance ofpolymer concrete samples to this severe wear test was far superior to that ofasphalt and Portland cement concrete reference surfaces.

Two problems were encountered towards the end of the program whichremain unresolved: (a) accelerated storage stability tests at 60'C pointed to

a stability problem with mixtures of mercaptan resins and the tertiary amineaccelerator, Capcure EH-30, and (b) the flexural strength tests were quite lowunder the majority of application conditions encountered.

As a follow-up of the present program, it is recommended that the* stability problem be more carefully examined to determine:

(1) Its severity under lower temperature conditions.

43

(2) The possibility of substituting alternative tertiary amhnes orother accelerators for Capcure EH-30.

(3) The effect of substituting alternative diluents for the Mer-captate Q-43 Ester (either reactive or nonreactive types).

(4) The possibility of adding stabilizers to the resin/acceleratormix.

The results of the flexural strength tests demonstrated that the lowvalues obtained within one hour of application are attributable to the highexotherm of the curing reaction which requires a long cool-down period. Whenlet to cool down slowly or when quenched with ice water just prior to testing,flexural strength values of about 1500 psi were obtained with beams preparedunder dry conditions and values of 1200 psi with beams prepared with wet ag-gregates.

Two possible approaches may be considered to address the problem ofthe exotherm and the heat content of the repair cap shortly after preparation:

(1) Physical cooling of the repaired cap with dry ice or othersuitable coolant; this is technically a simple matter and would probably besuccessful, but must be examined from a logistics standpoint.

(2) Reformulation of the epoxy-curing system to minimize the exo-therm problem and to build rigidity into the polymer-concrete within 1 hour ofits preparation.

A follow-up effort is recommended to examine the feasibility of these appro-aches. A discussion of the theoretical considerations for achieving the ne-cessary chemical modifications is provided in the next section.

7.1 Discussion of Theoretical Considerations

It has been shown earlier that using a mercaptan type curing agent,bisphenol-A based epoxy resins can be cured within a half hour, as desired.However the excess accumulation of heat takes time to dissipate, and theresin-aggregate mix is kept hot and relatively soft. This makes the runwayunsuitable for landing. It would be desirable, therefore, not only to achievethe necessary state of cure of the epoxy compound but to get it j.n reasonablyrigid condition within the scheduled time.

The accumulation of heat is brought about by:

(1) A high heat of polymerization (AHp).(2) Increased rate ef polymerization (Rp).(3) Slow rate of heat transfer, (Rt), through the organic medium.

7.1.1 Rate of Polymerization (Rp)

It has to be kept in mind that the high rate of polymerization(curing) is essential to achieve a good state of cure in a qhort time. Thesharp rise 1.n temperature at the beginning acceleraten the process. 'ut once

q4

the state of cure is achieved, the heat should be dissipated faster. In otherwords, a short time-temperature peak is ideal for this process. Alternativelyan optimum temperature for the exotherm may be worked out where the temperaturerise is moderate but high enough to carry out the polymerization and low enoughto ensure rigidity of the finished product. This can be achieved by selectingan appropriate resin system.

The mechanism of epoxide reaction with amines, phenols (oxey andthio), anhydrides, etc., has been well described by Lee and Neville *. Al-though the rate of reaction depends on the molecular structure of the epoxideand the curing agent, it is mainly controlled by the accesibility or stericfactor. Table 8 lists the time required for gelling of diglycidyl ether ofbisphenol A (DGEP) by a number of substituted amines, and it is clear that thecuring agent having the smallest substituent (Benzyl dimethyl aniline) reactsfaster than curing agents with large substituents. This is obvious from thescheme of the reaction given below:

0 '+ OHH -- o OH -

CC+RN-- C-0C.Z - C -C + RO3

6- 8+ R3NR3N

0 0

c- \c.--> C-c

RO OR

Thus, the tertiary amine has to get adequate space to approach C atom I toinitiate the reaction. The accessibility of the hydrogen donor H..OR isequally important. Since primary and secondary amines, after donating theirhydrogen in successive steps become converted to tertiary amines, the mechanismof their reaction remains quite similar to that of the tertiary amine. It isimportant to note that the availability of active hydrogen is part of the cur-ing process. The reactivity of the hydroxylic group and the crowding in itsvicinity determine the rate of the reaction. Similar correlations between therate and the structure of other curing agents, such as acids and phenols, havebeen established. The reactivity of the epoxy resin is also important. Thisagain is basically determined by the accessibility of the epoxy group and theelectronic-nature of the epoxy oxygen. Given good accessibility, electron at-tracting substituents such as methylene and vinyl improve the rate of reactionwith acid type electrophilic curing agents. On the other hand, in the presenceof a base catalyst glycidyl ether will act faster compared to epoxidizedcyclohexene or vinyl cyclohexene. Finally, the rate of reaction also dependson the number of epoxy groups in a resin molecule, and this can be variedwidely. It may therefore be possible to control the rate of the reaction byproper choice of epoxides and curing agents.

* Lee, H., Nelville, K. "Handbook of Epoxy Resins", McGraw-Hill, inc., 1967.

45

TABLE 8. EFFECT OF SUBSTITUENTS ON TERTIARY AMINE ON REACTIVITYOF DGEBA

Amine Gel time, min

CHi

Benzyl dimethyl amine C 5.3

CH33

C HBenzyl diethyl amine PV H 12H-N 7300.0

2 5

CRH12 5Triethyl amine C H N 11.2

2 51

C25

Tripropyl amine C 37 29.0HC-N

C37C49

Tribuhyl amine HgC -N 33.094'9

Dimethyl ethanolamine HO-C2 H 4H 3 4.3

CH3

CH 5Diethyl ethanolamine HO-C H -N 17.2

C2 H5

46

7.1.2 Heat oE Polymerization (AHP)

The heat of polymerization depends on the chemical nature of thecuring agent. Some of the measured values of AHp with diglycidyl ether ofbisphenol A are given in Table 9. It would appear that the maximum heat isgenerated when curing is carried out with the primary amine and the minimum inthe case of a combination of a highly substituted tertiary amine and a carb-oxylic acid. Hence the heat evolved and accumulated during polymerization canbe adjusted by choosing proper catalysts or combinations of catalysts.

7.1.3 Temperature Profile

The exotherm, which is a measure of the highest temperatures reachedduring curing of an epoxy compound, is also a function of the temperature ofmixing. The exotherm profile of DGEBA and DMP salt is shown in Lee andNeville. * It will be seen that whereas mixing at 230 C and curing at 460 Cgive highest exotherm it lasts only for a short time compared to the one whichis cured at 300 C. On the other hand, mixing at higher temperatures and curingat 230 C make the exotherm peak around 800 C and cool-down quickly. Thussuitable combinations of mixing and curing temperatures can be used to controlthe peak temperature and the life of the exotherm.

7.1.4 Heat Transfer

The heat transfer coefficient of the compound depends on the natureof the resin, the state of aggregation, and the macroscopic structure. Usuallymetallic and ceramic materials transfer heat much more quickly than the organiccompounds. A number of fillers can be used with epoxy resins to assist thetransfer of heat from the finished product. Again fine particles, because oftheir greater surface area, may transfer heat better than coarse particles, andmetallic fibers in particular may be well suited for this purpose.

7.1.5 Temperature Rigidity

The temperature where the the polymeric product becomes rigid (glasstemperature) again depends on the rigidity of the backbone chain of the polymerand intermolecular force density. Epoxides with bulky groups on the side chainhave higher glass temperatures compared to the linear ones. Polymeric epox-ides, particularly the ones based on Novolacs, tend to have higher Tg. Al-though a finished epoxy product is a thermoset and the concept of glass tem-perature does not apply, the principles can still be used to design the basematerials and, by suitable choice of epoxy materials, products can be obtainedwhich are rigid at higher temperature.

Whereas epoxy resin weith mercaptan type curing agent has demonstratedthe possibility of repairing the runway within an half hour, the residual pro-blem of achieving rigidity within that time can be solved by adjusting themolecular and physical parameters as outlined above.

* Lee, R., Nelville, K., "Handbook of Epoxy Resons", McGraw-Hill, Inc., 1967.

47

z -4

rA r-41 4cc en CIA

tw

C4 w

E-44

41r4

O,4410 0

01P-4.0

'-4

r340f

f 448

APPENDIX A

SUMMARY OF TABLES OF EPOXY FORMULATIONS

49

.0 -.

-~U

4*

-.

a

S U

be be.0 .0.0 .0

N .0

.9 ~ - - -I

I *

'4

be

* U ~ Sbe be be U U .0 ~ A

.0 .0 .0 .0 N N N A.9 .9 N - - -

ii2* be

- N N N N N N N g~

beU N N N N N N N N Nj *

U *

AEli

U1044 N N N n N N N N N NN N N N N N N N N N S

2~ 0cc a a ! !;

Ij be

0* a a a *

2 ***

I IB

I I

20

C -.9

.9 .9 .9 C .4

be ... .

. SS N S ~ begN

NI I.

''SiM

2.0 0

4 *~5N

~

.3

LU..4UUU

- a Ube 1'

~ i. ~.'5

U-

~ I ~

'~jIN N N. NN

3 N~ N

2

a * a 0 ~- -- be A

N

4.&

50 IA

Ol- - - - -~ -

'AA A. .

19 0

* Ua

IL 'A a a A a

4 s as As -ts t

*a

2 1

- .9 .9 .9

~ 3 - - '- I,

g

'U~.I.- - II W~ ~ll-- .94' .9Pta - .9 .4 *tt . .4

~! 2 2.9 .~ ~. -t -

3 - I..a

iik& i~ ~

a.08

b *9 ~9 ~9t9 .9~

.9 1~ .4 *4 .4UI-

B SU. -

3 ~

2 0

- - *1B

- P -. I.* R a. a a

U a. a a - -

U 'aa* .4 8%4 -

.9.9 it2a. ~ :ii5 - .9 .9 - -* a -

.4.4 4... . no.., 4.

I ~ .4.4.9 .4~

4. 11111111:!

ilfl*I ~

~ sJ~1 4

-- -. 9 .9 .9.9.9.9=

52

#4

S : £ ~

I B I UU- a -

- ~iI I ~4.. g s

I~ U

~ £ £4, US #4 -.

~RE £ £ £ £ii

- I.I -, 4, *~ *fl 4~ 4, 04

I- 4.4 4.4 - 4.4 4.4 4.4 4...

U

U-u

I I.4 a a ma

I'll H- 555 iiA a -U 0 -S.* -

S. S..1 0 0 0 0

U.. aaa a a a aI ~~33 * b b 1!* * *

I -3 I 0 04' ~§ ~ a aa-

aaa a a a a .3u U31'*1

U ~I~!I II iiLI jjjj. ~I 1.1 U

I;k *~. ~* 4, .q - ii.1. .1.

4,4.4 4.4 .1.a U a

53

6;

a do4

v , ,

u01*1

'I19

.131

I R54ijh

. . .....

i . .4 I 4 t 4 11 1 4 i *

I-I

. . . . . . . a i

.4 .4 .4 . .4 . t4

Igi

a' 4

11U

-- •.. - ft f4 tU

A.. i 111 1 *

ft

* U U U *

II

g .* N

' ii

a. n ~ - V *I...ij ~ - - - -

Ga 3G~* 3~

VII * I'-~ 11 1 II U I'~E ~ u 1 U

~44,~* -. ~.#4Ah

3 3 1* 4, 2

U* .14

I ~E

~* 140

ii1 4,

-n 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4,

B* I.'1. 2 * * 0 - * 0 6 6 03 ~~~4,

.. aaaaa a a aaaa a~* . 6 6 S I * * * S S I

14 14 14 14 14 14 14* U S U

U U31a1I 5. 5I ~ ~g~2a2~ 22 SSUS ft

I III1IIiiIi ii ii1 11111

a!UI~U ~ 6U *0 U U U

I 111111 £ ii IIII~ I I

* ~ Ia -- ~ 5 5

57

4~ - -' -.

*ijm - - -. -.Ib~R .4II.~u .~ -

31111 1 3

I I*4.4 - -A N

1111 1 1I ~ -

U.a.-. - .4 4n 4n 4n en

44U *4 *4 *4 .4 .4 .4

k-u .4

* a

~ ~ 6

UII~ .4* - *

i.~ - -4

S. U

C-c. -IUtS. -o a.. S. 0

S. * U S. .4p 5 I 6 * * * * S S..

- *d 54 ** *j ~5 5

U;;~ A' A' U 'U

1!~ 2 ~ ~

d 0

3 N

* - U -

3 ~-- -~ -3 aI Ii Ii I~ ! ~

3 * *' U a *

34 IllS U ~a a ~ lj

p IiiIr5~58

177849A

-k m4 m-~~db .4.,.4.,.4.

UI

go z I I

92Z

Ii- 59

- U* -~ -I~i: - ffl U U

S oft- ~

-1 *2 I~ 11 ; I I II

ft .~, ,. ft U U

~. EU U I U Us3.! I I I .~ I ~ I 1 11

-, - ft fltfl

s u~ u:

U U-

-3U. -~ fi

C, Cd - Cd Cd Cd Cd Cd Nfl4

1U ft N ft ft ft ft N ft 95-.

a2 -~

I E~ - vs vs Cd Cd

ft ft ft ft ft ft ft ft ftI- U5.-

a 8! 0 00 0 0 0 8

R

** U 5. U U 0 U U * *- C U 4U U U U U I

~' 3 ~ .1 U

a. .~t

U88888 ~ a 0

ii ~A.1 5 A A A A. U 0 0 U U

5. U - U U 5. 3 I. - -5. -3 ~ p aa K j. ~. . U a. Uaa~ 33 3 a a a U

A -. U

222 22 2 2 2 - ~ - S

*gg ~a U K

k&k && S Si 1 Cd ft* U U U

&I! ~* ft Cd ' ~ * A U U

60

- -U , . '

.t % % I I % II . I

ou0Oa1I I u a I ! - 04iI i Nbi i i 04 f4 -

'61

::j.

** I

.I - I

66

> t, ' "

. . a

I I I

ii

N R1N-- i

• it i ,i ,i i f i

* i " ' ' , - - -

,f f f fif i I if if if , i -

if'f if ii~,if f

* bUI . u LI t

if. ~ 6

'a -I.. - - - -~ - ii- - C, _ C, C,iC!I - - - - - - - n~ ~ - -%-

C,4.hh CM CI

~ .. '. ~C. C, -. C- 4 * C, R C * 5 CC ~

I I-~ I, 111 I I U II~ ~ N ~j ~ -'-C 4 -~ 4 '0

0 ~* IC~* UU~;3 U .1~ 1IC~~1 I I I *o CI

- - '0 @1 C.. 4 *

- , -8M

~- II

* ~ ~ ~a

I-

5 C, C, ff1 C, C, C, C, C, C, N C, C, 4, ~ C,a ~*4 CM CM C'4 CM N CI CM CM CM CM C4 CICIa I@0 4 4 4- - - - - a a a * 0 0~aaaa a a a a a a a a a

- - 6 66 6 6 6S * * * 6 ~ - ~ U S S

CI CI

-,.wu HiH~ ~

C '' * 0. 0.SS 88 NN

N

a a *~ -5 III Ill181 7~s;;;J;jj;j~ JjjjjC-C, NC, C-C, C-C, NC, NC,C.

. --CC,

~ N 319 419 3 N CN 3~ 'C

I ~ 3!z:::z::..V ~ 1 - ~ 3

U -- CIsU; UU * *~CI

~ .~ ~

~ A CM

~61.

* 5 5 IIn0

I~1412 4 4~IU 4~4 4~4 6

U

i~ Ij~

~44 I

U- S.4.4 44

~iU ~

S

- S-

I.4., .6 0 0

~ag -II~~S 4. 44

N

§ a

~ ~liii''.4 ~ijIa1I ~1ij~J

LI&

~

S.

65

J~irif - - -I, N N N N N - ft ftaa ~ ~ ~

U C N N I I

~ III~~- NI . nil - U~.u 11113 1

I N

6 _ N

ai~ liii Na2 ~ ~ ~

a'1.

* N N N N N N N N N N N NS kU ft N N N N N ft ft ft ft N N

a

*I N ft N ft Nft N ~ ~ ~N fl N N N N N N N N N N NN H.

a - -B

C. U

a St

- - - 0 553 0 0 0I. 22222 2 2 2 2 ~1 I U *.

I' aI. C S ~

I I C.C. I. C. ~9 1 1 ~N a -22 22 ~ U -

* 0N -.IiI~, E~ L~I: .*

5 U 5 - - -

* * * I* a 0 l.a65 65aH. e~I Ilil 21U A

jg ewe em . .~ U U 3 3

I!* - ft N - N -

a S A

66

IA; - - - w~oo '- .-a a .9 a - a .9 w~ . -. -I - ' 1 3 3

4 .9 .9

I ~ U ~ U- I - I

I. - * 04 a ~ . m *4 -.1Ia I

~ Iji I= W 04.9 g3..3

~ * a i~ * S ~2.LI.L1U

£ £ £ £ £ £ £ £ £ £ £

~ ~Ji~ .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9 .9

a a a a a a a a a a a a

I...~b I~I3 ~ ~ .9 .9 04 04 .9 .9 04 04 .9 .9 qiEA~ a a a a a a a a a a a a

I.9 .9 .9* U= -

'U- 9

* v.3.1 .3

@ a - 2 2 0 2

B a a a a a a a a a a a a iia -jiji ~ U-S. U ** 66606

9 , . .a a a* uZKI~KKIKK K K K2 a UI. U -

.9.9 .9.9

S 32822 1~3.'aI U;

8 1 3.5,2 .5,

a.,

U.9 *~.9- M . . S.J

a-AUi; : - -sI UIR ~ Ia *a *~ - U

j. j fl~ .qc- .oa .. a j a a* - a .9 4

--------------------------------------------------------------------* a U

67

J e~a - - - - -~N N N N N

4 a * 4 N 4N - 4 P - N I I I

tarn ~. a'. ..

a a3 U II I N~ ~

Si Si S.

II 4, 4. 46

SiN

.u .~

N -I~1 4 I a

£ I £ I- "aK

3 -- 4.

9 tI. N N Si N N N N N.~4.U N N N N N N N

S

I... N N N N N N NN N N N N N N

5L NI a.. .0a

N 5 N N ItS 2 22 ~ 0

3 22 22 22 2 2 2I * UC

a. ; ai~ I. Ii 0a

4N NS S'4 4 N*"Q* * C~ 59 5 NO NUNq NUN4 C

NNN NN C -I S.'.3>3 A A~3 AZ 1a ~ aI. ~~ a ~a a:

~ 4a4 P.54 P.5'S S..'S ~5'4 N IN ON *N ON ON ON ON 4N 2 N

SN 4N OPS ON ON ON ON N

S. S

3I .. apSoS -~ .85 13ii ~ IiI IIII liii

- .5' - *. iiPS a

68

.. 04 I 4- *IS

* .44..04I I 0h "."U~~~ .0. -4.00* U 0404 0

.0 0 4 - 4

* e

Oi *

z 'A

4. .04 it -i -I - 0 M.4 .4 4 - - - .4 I

I I Ia

A4 24 A I A a A !A.

I A' * . V' 1 H f H1 A gal

3 " .4 .4 fi ( 7

4 l, , 4, 4 .

! .I7O

I.

I

*

- - 'Sn SnU * * * Sn Sn Sn'Sn - - .- - - .- - U

Sn Sn f - 5-,,Sn Sn C CC Sn 0 Snjj3 7Y '1' -- - I --- ~ 3 ~ u .. A ~ J~ Sn'S Sn '. 'S Sna

~ ~b Sn Sn Sn Sn Sn- -3

SnSn Sn SnSn Sn 'S 'S ~' "S S ~ - - Sn Sn - *~4 Sn

U4 b.

h.a~=

Ed ~aa a2~U

U 31-S

U.-. Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn SnUU Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn SnEd Sa

U

Sn Sn Sn Sn Sn

Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn C

aMa Sn Sn Sn Sn Sn Sn Sn Sn Sn

£ 0 0 a 0 0 a a a a a a a aI R3 ; aI~I, SI.~.L.L' ~J... * smul 'Sun umu -uS.

S P.* 0 6 S S 56 5 56 * US-. U U US. S. U I. U U* S 4 44 4 3 3 4 4 3 3Sn U U U YU U V U

Sn L I. U U

Ak A s~ s ~ I0 5.

3 ~ I a a~ ~ a' ~ ii* Sn*r4

5SnSn Sn Snj CC Ci Sn SII9~U I.~ U6S.~ - . ~*** .. *~ ~I:IRZ!I~1 3UtI~~t **tI~ ii 13 13

4 ~*~)~ik~ J~iJai~ IJafi JI~*I* -~

a

iu~~au uauu ~ ~ 'IiI' ~ ! Iiift fl ~Sn C ~Sn Sn Sn -

.9.- 4 'S -. 5-Sn

S

i

)1

AJ

- 4 . .4 .4 .i8A 2

.- , 4 - .4

A - -U 44

- I~ I

0

-I--Ill I

ill*S...! a a a

S. - - - .4d - - - ~ A

U

U ' -, -' -s s.s .s -s U~* 440 - .4 .4 .4 .4 .4 .4 .4

S Sk* ~u.III ~

to I..q S. U

U UU

S. .4 .4 ~ ~ ii.0 -

iii ~I

I U

31-

~

:23.6.

I ~- I *uu

.4 P4 ~ illi~ * ~ ~ * * - -a S

3

- - en en en en enIII .

* ,~ ,..

0

en - en

en N en- - - - N - - N

2.1 en-g I.... N4c3~ I - I I Ii

I-~i ~ iki

U'

-I - 2~A~0 en en en en en en en en en en en en enN N N N N N N N N N N

Ii. I I

-U.0en en en en en en en en en en en en enN N N N N N N N N N N N N 0

N hN 1

IB- S S -

U N 0-I a -I III I I I

N I U S

4* en en

a a 0 0 0 en en ~ en en en 0 N

U U#Z. U.S 5.5 5 5 5 5 5U ~ U :5~ ~ N N N - N

IJ U U U IS U IS Uijiji - - - - - - - - Aiii I S I IiM enen enen

UU N

I3 - * * Se

2 2 2 2 2 ~ ~.u ~ - - ~ I i

& iLi~~ aI

* ~ ~ - -JI.A±.L I&.~S ~ I,.

4 4544 *~* @5 Se

7'

A -.-

-a -

a i

TABLE A-24. LEWIS ACID CURES OF CYCLOALIPHATIC EPOXIDES

Cure RateFormulation Epoxy Resin Curing Agent phr of Casting

73-1 Araldite CY179 Resicure 30 10 (a)

73-2 Araldite CY179 Resicure 30 10 10 min (b)

81-1 Araldite CY179 Dibutyltin 10 Did not cureDilaurate

81-2 Araldite CY183 Dibutyltin 10 Did not cureDilaurate

7 5-1(c) Araldite CY179 NI4A 127.2 Did not cureResicure 30 10

7-() Araldite CY179 NMA 127.2(a

Resicure 30 10

80-1 Araldite CY183 NMA 115.8 Did not cu~reResicure 30 10

80-2 Araldite CY183 NMA 115.8 Did not cureResicure 30

(a) Set up in less than 25 seconds; could not be cast.

(b) Strong exotherm; resin charred.

(c) CY179 and NHA premixed and Resicure 30 added.

(d) NHA and Resicure 30 premixed and CY179 added.

76

APPENDIX B

SELECTED POLYMER CONCRETE LITERATURE REVIEW

A comprehensive literature review of polymer concretes or ofrunway repair methods was not within the scope of this program. but a numberof relevant reports were reviewed to give perspective to the presentstudy.

The problem of rapid runway repair has been addressed by variousinvestigators for 10 years with varying results and with differing con-clusions being drawn. The approaches to the problem varied, since theobjectives were not uniform: some sought repair methods only for temperaturesabove freezing, others limited investigation to dried aggregate only, whilethe durability of the rapid repair was of importance to some, not to others.In a general sense, there appears to be a preference throughout for polyesterresin binders for polymer concretes. Some of the salient features andreported results of a number of these studies are reported below.

1. Leitheiser, R. H., R. J. Hellmer, and E. T. Clocker; Technical ReportAFAPL-TR-67-146. The objective of 300 psi flexural and 1000psi compressive strength was exceeded using polyester-styrene emulsions.Flexural strengths of 500-2500 psi were achieved depending uponwater content. However, these measurements were typically made manyhours after the resin was applied. High temperatures gave poorresults because of too rapid gelation. The use of aggregate (unspeci-fied) was unsuccessful since it did not bond to the resin even when asilane coupling agent was added. Although not stated in the report,it is obvious that an aqueous emulsion is of no possible use attemperatures below 320 F because of freezing.

2. Hickman. B. R., et al.; Technical Report AFML-TR-66-141.

A wide variety of approaches to the rapid repair problem are con-

sidered in this report, including:

a. Graded density polyurethane foam applied over a base to a depth

of three feet. A compressive strength of 440 psi was achieved,but no flexural strengths are reported. An isocyanate con-taining formulation is clearly unsuitable for wet aggregate(no aggregate was used here) due to its reactivity with water.

b. Molten sulfur with gravel provided a compressive strength of1500-2100 psi, but the temperature required for penetration ofthe aggregate proved to be too high for practicality.

c. Several inorganic soil binders were examined including sodiumsilicate-calcium chloride, but the product strength was poor andthe method of application too complicated.

77

d. Similarly, calcium acrylate-soil mixtures were evaluated, butlow compressive strength, slow cures, and water sensitivitymade this approach unattractive.

e. Epoxy resins were evaluated, premixed with aggregates of varioussorts of which sand proved to be best; stone was not evaluatedas an aggregate. Flexural strengths up to 3750 psi were obtainedusing Epon 828, triethylenetetramine and boron trifluoride-diethyl amine complex blended with sand, then poured. However,only Epon 812 bonded to wet sand, and the strength of thisproduct was poor. In spite of deficiencies, this work lendsencouragement to the epoxy resin approach, since better results wereachieved under certain conditions than with the other systemsstudied.

3. Nielson, J. P. and V. Cassino; AFCEC-TR-75-25. An objective ofthis study was attainment of a flexural strength of 1200 psi. Effectsof temperature and wetness were not studied. Liquid resins were appliedto a 10-inch bed of gravel on a sand base, but the 10-inch depth was notfilled so that gravel particles were simply coated. A variety of binderswere evaluated. Objections to the polyester-type binder includedcooling by evaporation of styrene and residual tack. These objectionsdid not apply to the epoxy resins. However, flexural strengthsmeasured several hours after the test beams were formed were only abouthalf that desired using 3/4-inch gravel.

4. McClain, R. R., in "Polymers in Concrete", Second International Congresson Polymers in Concrete; University of Texas, Austin (1978). This workis not directly applicable to runway repair but does indicate thatimproved properties can be achieved when epoxy resins are blended intocement concrete (wet) mixes for use in patching bridge decks.Strength is slow to develop, but the concrete is significantlystrengthened by using an amine-cured epoxy in the mix.

5. Peasley, J. A., Air Force Aero Propulsion Laboratory TR-65-120.The objectives of this study were to achieve satisfactory runway repairwithin 4 hours to' support wheel loads of 25,000 pounds. Applicationshould be possible within the O'° to 1100 F range (minimum) and from dryweather to rainfall. The report recommends polyester with dry aggre-gate. Epoxies are reviewed as imposing too exacting demands to bepractical. Silica aggregate was used, but no coupling agents wereevaluated.

6. American Concrete Institute, "Polymers in Concrete", State-of-the-Art Report, 1977. This report is not directly applicable to thecurrent work and is rather general in nature. Polystyrene is citedas the most commonly used organic material, but furan, epoxy and poly-ester compositions are also noted. The use of coupling agents isnoted, and only silanes are identified. These produce the best strengthenhancement when the aggregate is pretreated rather than when thecoupler is blended wit the uonorier.

78

Z*

7. Keeton, J. R. and R. L. Alumbaugh, U.S. Navy Civil Engineering

Laboratories, TN-1479 (1977). This report deals with polymer additionsfor concrete rather than the use of polymers as the concrete binder.

Only two resins were examined, an epoxy and a Saran latex. The epoxymodified concrete developed flexural strength more rapidly than theSaran and had greater ultimate (one-year) strength as well. The useof coupling agents was not mentioned.

8. McNerney, M.T., Air Force Civil and Environmental Engineering Develop-ment Office. TR-78-10. This study was confined to the use ofacrylic monomers polymerized with gravel aggregate. Any significantamount of moisture reduced the strength of the finished product;rainfall precluded use of this system. Use of a silane coupling agent(Dow: DC-7-3060) increased the compressive but not the flexural strengthof the samples. The polymer concrete proved to be significantly strongerthan ordinary cement concrete.

9. Setser, W. G., et al., Air Force Aero Propulsion Laboratory TR-67-165.This study is notable for the large number of polymer systems

which were screened experimentally: 170 polyester and vinyl estermaterials and 61 epoxy resin formulations. The objectives include:cure in 30 minutes or less, rapid cure at -5 OF and a working rangeof -5 oto 1100 F, curing in the presence of water and high tensilestrength when hot so that the repair would be functional during thelong cooling period. Epoxy systems were disqualified early in theprogram because of high viscosity and high cost. Of the organic

systems only one polyester and one vinyl ester formulation wereselected for final testing. Fine particle filler was used to enhance

product strength and reduce the exotherm. Addition of a silanecoupling agent enhanced the flexural strength of silica filled resinformulations. The overall approach was directed more toward pre-paration of surfacing formulations than the use of a resin as abinder for aggregate. The final selection was an inorganic cement whichproduced a flexural strength of 600 psi and compressive strength of3600 to 4000 psi within 30 minutes of pouring. The success of thisapproach in use was greatly dependent on the preparation of thebase.

10. Smith, A., Air Force Civil and Environmental Engineering DevelopmentOffice TR-77-53. The goals of this study included a 30-minute cure in the 35 oto 125 OF range using polymer poured on wetaggregate so as to fill the void space. Laminac 4128 polyester mostclosely met these requirements. Epoxy resins and urethanes were alsoevaluated. However, fiberglas reinforcement was required and dryaggregate was specified. With these limitations the resin at 500 Fcould be applied to aggregate in the range -22 0 to 122 OF and producea product with a flexural strength of 815 psi in 20 minutes, 1057

psi in 45 minutes.

79

I _________________

APPENDIX C

PRICE, HANDLING, AND SAFETYINFORMATION OF FORMULATION COMPONENTS

All resins and additives used in the two formulations selectedfor full evaluation (116-2 and 122-1) are commercially available. Descrip-tions of the chemical characteristics are given in the section on ResinFormulation and Screening. Following are the price, safety, and handlinginformation provided by the manufactuers of the products used in this study.It must be borne in mind, however, that equivalent products may be avail-able from more than one supplier.

This information is only provided as a general guideline. Forcomplete information, the manufacturers' data sheets and recommendationsshould be consulted.

Capcure EH-30

2,4,6-Tri (dimethylaminomethyl) phenol

Diamond Shamrock CorporationProcess Chemicals DivisionP.O. Box 2386 RMorristown, N.J. 07960

Specific Gravity 0.97Flash Point, Open Cup 3000 F

Price: 2000-5999 lb $1.49/lb6000-29,999 lb $1.37/lb30,000 lb to TL $1.35/lb

-Storaste and Handlinst: Capcure EH-30 is shipped in lined openleadsteel drums. Temperature changes have no adverse effect on the propertiesof the product.

Safet, Information. The acute oral LD50 is 1.7 g/kg body weightin rabbits. Slightly toxic through skin absorption and inhalation but nota sensitizer. Potentially can cause severe burns on contact with skin andmay act as a corrosive on prolonged contact with skin. Prolonged over-

exposure can produce irritation of skin and mucous membranes, eye injury,

delayed lung injury, and chemical pneumonia.

Splash goggles, protective gloves, and NIOSH-approved respiratorsfor organic vapors are recommended.

80

Caocure 3-800

Mercaptan terminated liquid polymer

Diamond Shamrock CorporationProcess Chemicals DivisionP.O. Box 2386 R

Morristown, New Jersey 07960

Specific Gravity 1.15

Price: 2000-5999 lb $2.32/lb6000-29,900 lb $2.21/lb

30,000 lb to TL $2.13/lb

Storage and Handling: Capcure 3-800 is shipped in 55 gallon lined

tighthead steel .drums. Neither chemical nor physical properties of this

product are adversely affected by temperature changes.

Safety Information: The oral LD50 is over 2.6 g/kg body weightin rabbits. The aermal Lu50 is over 10.6 g/kg body weight in rabbits.Mildly irritating to eyes and skin. On ingestion, call physician immediately.

On contact with eyes, wash with water and call a physician. On contact with

skin, wash with soap and water.

Splash goggles, protective gloves, and NIOSH-approved respirator

for organic vapors are recommended.

Epon 828

Dilycidyl ether of bisphenol-A type epoxy resin

Shell Chemical CompanyP.O. Box 2463Houston, Texas 77001

Specific Gravity 1.16

Price: Truck load drums $0.98/lb(delivered)

Tank truck (delivered) $0.93/lb

Storage and Handling: Shipped in 55 gallon steel drums.

SSay Information:

LD5 0 oral - 11.4 g/kg body weight in rats- 19.8 g/kg body weight in rabbits- 15.6 g/kg body weight in mice

81

LD50 derual - >20 g/kg body weight in rabbits

Contains 1-2 ppm epichlorohydrin, a cancer suspect agent.Irritating to eyes and primary irritation to skin on prolonged or repeatedcontact.

Eye protection, rubber gloves protective clothing and respira-tors are recommended to prevent body contact and inhalation.

Hteloxy 69

Resorcinol diglycidyl ether

Wilmington Chemical CorporationPyles Lane. Hamilton ParkWilmington, Delaware 19899

Specific gravity 1.20-1.22Flash Point, Open Cup Exceeds 3000 F

Price: truck load, 55-gallon $3.25/lb

drums

Storage and Handling: Heloxy 69 may crystallize on standing.If this should occur, warming the resin will melt it. No change in pro-perties occurs during crystallization and melting. Storage is recommendedin tightly sealed containers in a dry place at ambient temperatures.

Safety Information: Short-term exposure to vapors may cause eye.nose, and throat irritation. Prolonged and/or repeated contact may causesensitization and dermatitis. Avoid fumes from material. If hot, over180 F, wear a respiration mask for organic vapors. Eye protection withside shields and rubber gloves are recommended.

In case of skin contact, wash contaminated skin with soap andwater. In the event of eye contact, flush well with water and consult aphysician. If ingested, induce vomiting if the patient is conscious andthen call a physician.

Mercaptate q-43 Ester

Pentaerythritol tetrakis (mercapto propionate)

Cincinnati Milacron Chemicals, Inc.West Street

Reading, Ohio 45215

Specific Gravity 1.28Flash Point, Open Cup 4900 F

82

Price: 500-4000 lb $2.35/lb4,000-10,000 lb $2.30/lb10,000-24,000 lb (truck load) $2.25/lb

Storage andandling. Hercaptate Q-43 is shipped in polylinedsteel drums. Excessive contact with iron and non-ferrous metals shouldbe avoided.

Safety Information: The acute oral LD50 is 872 mg/kg body weightof male albino rats. Mercaptate is considered toxic but not highly toxicby oral ingestion. Exposure of the eyes and skins of rabbits indicatedthat Hercaptate Q-43 ester is considered nonirritating.

Safety glasses and protective gloves are recommended.

THPTA

Trimethylolpropane triacrylate

Celanese Chemical Company1211 Avenue of the AmericasNew York, New York 10036

Specific Gravity 1.1Flash Point, Open Cup 2800 F

Price: Truck load, 30,000 lb $1.15/lbTank wagon, 40,000 lb $1.10/lb

Safety Information. Acute oral LD5 0 is 500 to 5000 mg/kg bodyweight in rats. Acute dermal LD5 0 is 2000 to 20,000 mg/kg body weight inrabbits. Harmful if swallowed, inhaled or absorbed through the skin.Mild skin irritation at 72 hours. On contact with eyes, corneal opacityreversible within 7 days.

In case of contact, immediately flush eyes with water for atleast 15 minutes. Call a physician. Wash clothing before reuse.

83

(The reverse of this page is blank)

mo

INITIAL DISTRIBUTION

DTIC-DDA 2 2HQ AFSC/DLWM 1HQ AFSC/SDNE 1HQ AFSC/DEE 1HQ USAFE/EUROPS (DEXD) 1HQ USAFE/DE 1HQ USAFE/DEX 1293rd Zngr Combat Battalion 1AFATL/DLJK 1AD/IN 1USAPTAWC/THLA 1AFATL/DLODL (Tech Library) 1AFTEC/DET2 1EOARD/LNS 5SHAPE TECHNICAL CENTER 1HQ PACAF/DE 1AFW.L/DEA 1HQ TAC/DRP 1HQ TAC/DE 1HQ TAC/DRPS 1HQ TAC/DEPX 1AUL/LSE 71-249 2HQ SAC/DE 1US Navy Civil Engr Lab 2HQ ATC/DEE 1HQ MACIDE 1HQ TAC/DEE 1HQ AFESC/DEMP 1HQ AFESC/DEO 1HQ AFESC/TST 1USAF Academy/DFEM 1USAE WESGF 2HQ USAF/LEEX 1HQ USAF/LEYM 1HQ USAF/EDPX 1AFWAL/MMXE 2AFWAL/FIEM 1HQ AFLC/DE4G 1AFIT/DET 1HQ AFLC/DE 1AFIT/LDEIMQ AAFCE (Log Plans) 1TAC ZEIST 1550 CE CFE HQ 1LuftflottenKommando A 3 V 1Luftwaffenpionierlehr Kompanie 1Stat-Major 1Etat-Major Force aerienne 1Defense Equipment Staff 1HQ AFESC/RDCR 15AFIT, Tech Library 1MVEE(C) 1AFWAL/MLBP 1Procurement Executive 1

85

(The reverse of this page is blank)